Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
IMMUNOLOGY, HEALTH, AND DISEASE Directional Selection for Specific Sheep Cell Antibody Responses Affects Natural Rabbit Agglutinins of Chickens P. F. Cotter,1 J. Ayoub, and H. K. Parmentier Framingham State College, Framingham, Massachusetts 01701-9101; and Wageningen Institute of Animal Sciences, Wageningen University, 6700 AH Wageningen, The Netherlands intermediate, and line L was lowest. The correlation between SRBC and RRBC titers was 0.43 (P = 0.0). Females had higher titers than males, but the difference was only significant for the SRBC antibody (P = 0.028). Qualitative changes in anti-Gal accompanied SRBC selection. Rabbit agglutinins of 4 types were recognizable: classic, granular, annular, and one negative or very weak reaction. The score type means in line L were highest, in the control line were intermediate, and in line H were lowest, suggesting avidity differences now exist among these lines. The results show integration of natural and acquired immune systems because selection for one temporarily affected the other. Given the importance of anti-Gal in primates, our results should stimulate further study of this antibody in poultry species. ABSTRACT Agglutination data from generations 8 through 19 indicate that bidirectional selection for specific SRBC antibody responses was successful in a line cross of ISA × Warren medium heavy layers. After 11 generations titers of the high SRBC selected line (H line) were nearly 1:32,000; those of the low SRBC selected line (L line) were less than 1:2, but titers of the randombred control line remained stable at 1:32. Directional SRBC selection also affected levels of a naturally occurring rabbit cell agglutinating antibody (RRBC), presumably the avian form of α-galactose antibody (anti-Gal). This indirect response was biphasic and opposite in direction to the SRBC responses through generation 14 after which anti-Gal titers of all 3 lines increased. At generation 19, line H had the highest agglutinin titers; of both types, control line was (Key words: natural rabbit antibody, anti-α-galactose antibody, sheep red blood cells, directional selection) 2005 Poultry Science 84:220–225 diseases (Biozzi et al., 1979). Those observations prompted several groups to undertake similar experiments with chickens. The collective results have shown that acquired sheep agglutinin levels respond to selection in either direction (reviewed in Pinard-van der Laan et al., 1998). Selection for one antibody can also affect the response to other antigens. As a recent example, Yunis et al. (2002) found that chickens selected for Escherichia coli vaccine titers had correlated responses to Newcastle disease virus (NDV) and infectious bursal disease virus (IBDV). It would be of interest to determine whether selection for an acquired response also affects natural antibodies. Because rabbit agglutinins exist at high levels in several avian species (Cotter, 1998) they may provide a convenient test system to study interactions between acquired and natural antibody. Previously, higher levels of natural antibodies directed to exogenous and endogenous proteins were detected in the present H line (Parmentier et al., 2004). Furthermore, adoptively transferred natural antibodies enhanced subsequent specific antibody responses (Lam- INTRODUCTION The presence of natural antibodies to xenogenic erythrocytes in chickens was described by Bordet (1898) more than a century ago. Quantitative data obtained by Bailey (1923) established that in chickens, antibodies capable of agglutinating rabbit cells have the highest titers among 14 species tested. She also found that natural sheep cell agglutinins were typically at low levels in unsensitized chickens and only slightly above those for turtle, horse, and goat. The same author found that absorbing pooled chicken serum with guinea pig cells reduced the rabbit titers by 75%. The remaining 25% of the rabbit cell antibodies were refractory to further absorptions demonstrating their heterogeneity. Thus the rabbit agglutinin system of chickens is complex and composed of multiple types. Genetic selection for specific sheep cell antibody in mice is accompanied by greater resistance to some infectious 2005 Poultry Science Association, Inc. Received for publication April 10, 2004. Accepted for publication September 2, 2004. 1 To whom correspondence should be addressed: 39 Hathaway Cir. Arlington, MA 02476; e-mail: [email protected]; [email protected]. Abbreviation Key: anti-Gal = α-Gal antibody; H = high SRBC selected; L = low SRBC selected; RRBC = rabbit red blood cell. 220 ANTIBODY SELECTION AFFECTS NATURAL RABBIT AGGLUTININS mers et al., 2004), suggesting genetic and or functional linkage between natural and specific antibodies. Moreover the pathophysiology of protozoan and clostridia disease in primates is related to α-galactose antibody (antiGal; Pothoulakis, 1999). These observations provide a further strengthening of the case for the study of chicken anti-Gal because both are important avian pathogens. The objectives of the current research were to demonstrate that directional selection for sheep agglutinins also affected titers of natural rabbit agglutinins and that this was accompanied by changes in the rabbit antibody quality as well. MATERIALS AND METHODS Sera Blood samples were obtained from wing veins during the course of a divergent selection experiment begun in 1986 (Van der Zijpp and Niewland, 1986). Twenty samples from both selected lines were available at generations 8, 11, 14, 17 and 19; n = 200. Twenty additional samples were available at generations 8 and 19 from the randombred control; n = 40. These samples were kept in frozen storage (−20°C) in the Netherlands and shipped by airfreight to the United States, where they were analyzed further. Experimental Lines In brief, the lines described here originated from an ISA × Warren cross of medium heavy layers. Lines high (H) and low (L) SRBC selected were subjected to upward and downward selection for SRBC agglutinins, respectively. In addition an unselected control line was also maintained. Selection was based on individual merit using plasma microtiter levels determined at d 37 of age (Pinard, et al. 1992). This date represented the fifth day postimmunization i.m. with 25% SRBC. Rabbit Cells Whole blood of domestic rabbits was collected from the ear vein into chilled Alsever’s anticoagulant solution and stored at 4°C for no more than 1 wk. Erythrocytes were separated by low speed centrifugation, and residual serum was removed by 3 washes with PBS. Packed cells were adjusted to 2% vol/vol prior to use. Agglutinin Testing Microtiters were performed in 96-well U-bottom plates.2 Sera were diluted with PBS so that column one 2 Catalog number UN 2-9064, Marsh Bio-Products, Rochester, NY. Sea-Kem type ME, catalog number 50015, FMC Inc., Rockland, ME. 4 Goat-anti chicken IgM, catalog number 643951, MP Biomedicals (formerly ICN) Costa Mesa, CA. 5 Release 14, Minitab Inc., Malvern, PA. 3 221 was at 1:10 followed by doubling dilutions thereafter. Ten microliters of (2%) rabbit cells was added to each well and mixed with the serum dilutions. Plates were incubated at room temperature for 2 to 4 h, after which the cells settled. They were resuspended by vigorous tapping of the plates and allowed to resettle. Agglutination was determined by examination of each well containing serum by mirrored fluorescent light. Wells not containing serum served as negative (cell) controls. The column number of last well showing clear-cut evidence of agglutination was recorded as the titer. Qualitative differences in rabbit red blood cell (RRBC) agglutinins were recognized and scored by type as follows: 1 classic (blanket), 2 granular (gritty), and 3 annular (ring). Type 4 reactions (buttons) were considered as nonagglutinated and scored as 0 for the purpose of calculating the line means. Modified Ouchterlony Double Diffusion Glass plates were flooded with 0.75% agarose3 to a depth of 0.75 mm. Serum or IgM specific antibody4 was added to well rows punched at a distance of 10 mm. Based on an earlier method of diffusion distance measurements (Cotter, 2000) the locations of the precipitates formed between the sample and antibody wells were used as a gauge of molecular weights. This is possible because sample well-precipitin distances are determined by diffusion coefficient (molecular weight) differences. Precipitate lengths were taken to represent quantity as these are determined by the amount of IgM in the sample that has entered the zone of equivalence (Crowle, 1973). Statistical Analysis Data were analyzed using the Minitab5 statistical package. Titer data were analyzed using the GLM and interaction-main effects plot procedures. Correlations were determined by matching individual SRBC titers (obtained in the Netherlands), and rabbit cell agglutinin data were determined in the United States. One-way analysis of variance was used for IgM diffusion distances and RRBC agglutination scores. RESULTS The results of divergent SRBC antibody selection and its effect on RRBC agglutinin levels are shown graphically in Figures 1 and 2. These results represent the mean titers for the respective erythrocyte over the course of the experimental run. A direct comparison of SRBC and RRBC titers requires alignment because of the initial 1:10 dilution used with the latter. This is accomplished by addition of 3.3 (log2 of 10) to the RRBC values. The response to SRBC selection in these lines has been presented in more detail elsewhere (Bovenhuis, et al., 2002). It is necessary to include some SRBC information here as well because those results are interconnected with the current data, and together they indicate a certain degree of integration of acquired and natural immunity. 222 COTTER ET AL. FIGURE 1. Titers of SRBC agglutinins, determined by doubling dilutions (log2) are shown for control (C) and lines selected for high (H) and low (L) SRBC over the course of divergent selection. Data for line C were available at generations 8 and 19 only. Generation means shown by Figure 1 indicate that at generation 8 the SRBC titers were already clearly separated among lines. The log2 titer of the control line was 4.95, near to the exact average of lines H (8.6) and L (2.8). The response to downward selection was immediate in line L, and titer means continued to decline through generation 19. The response to upward selection in line H appeared to exhibit a slight lag because the mean for generation 11 was slightly less than at generation 8 (8.3 vs. 8.6). This delay was only temporary as H line titers rose and reached 10.1 by generation 14. The upward response of line H continued through generation 17 when the mean titer reached 12.9. Titers of line H increased FIGURE 2. Titers of rabbit red blood cells (RRBC) agglutinins are shown for control (C), high SRBC selected (H), and low SRBC selected (L) lines over the course of SRBC selection. Sera were diluted to1:10 initially, followed by doubling dilutions (log2). Titers of RRBC are lower than SRBC titers by a factor of 3.3 (log2 10) due to the initial dilution. Means of lines H and L within a generation not sharing a common superscript are significantly different (P < 0.05). Data for line C were available at generations 8 and 19 only. further to nearly 15 by generation 19. Thus by the end of the experiment SRBC titers of the H line would be greater than 1:32,000, and L line titers would be less than 1:2 when expressed using conventional terminology. There appeared to be some directional asymmetry in the responses of lines H and L. This difference was reflected by differences in the values of coefficients of quadratic fits to the data, where H = 9.9 − 1.45x + 0.07x2, and L = 0.45 − 0.08x + 0.01x2. This result is not uncommon in bidirectional selection experiments. Such asymmetry is sometimes considered to arise as a consequence of the scale of measurement (Falconer, 1981). Here it was probably due to the initial SRBC titer of line L, which was 2.8 at generation 8, already being close to its ultimate lower bound of zero, preventing further downward selection. Conversely titers of the H line were not subject to a similar constraint, but ultimately they might be expected to reach a plateau. In contrast to the directional changes that occurred in selected lines over the course of the experiment, SRBC titers of the control line remained virtually unchanged. Their final level of 5.05 at generation 19 was scarcely higher than the initial level of 4.95 at generation 8. Thus when expressed conventionally control line titers remained at a constant 1:32 over the course of the experiment. Figure 2 shows the corresponding generation means for the rabbit agglutinin levels (RRBC) in each of the 3 lines. As was the case with SRBC titers, lines were already separated by generation 8 with control line titers at intermediate levels. The initial titer of line H was about 5.4 (corrected to account for the initial 1:10 serum dilution) and actually declined to about 4.6 by generation 14. Conversely rabbit agglutinins of line L increased from 3.7 at generation 8 to 4.2 by generation 14. After generation 14, RRBC titers of both selected lines rose, and this continued through generation 19 in line H, but line L appeared to plateau. Thus the RRBC antibody titers of the selected lines were characterized by biphasic response kinetics. Initially (through generation 14) the direction of the response was opposite to what was occurring as the result of directional (SRBC) selection. The second phase began after generation 14 during which time rabbit agglutinin levels of both selected lines rose and reached their highest levels at generation 19. Figure 2 also includes the RRBC levels of the control line. Data were available at generations 8 and 19 only, which was also the case for the SRBC titers. Unlike the unchanged SRBC titers, the rabbit titers for the control line increased from 4.6 to about 7.5 by generation 19. Expressed conventionally this would correspond to an initial titer of less than 1:32 and a final titer greater than 1:128 or an increase of about 3 log2 dilutions. Although the kinetics of indirect response to selection were different from those of the direct response at generation 19, the rank of the 3 lines was the same for both agglutinin types. Line H was highest, the control line was 223 ANTIBODY SELECTION AFFECTS NATURAL RABBIT AGGLUTININS TABLE 2. Mean location of IgM precipitate diffusion distances Location of IgM precipitate2 Line1 Gen 83 n Gen 19 n H Control L 6.1y 6.9x 6.0y 10 10 10 5.4y 5.9x 5.4y 174 20 174 x,y Means in the same column sharing no common superscript differ significantly (P < 0.05). 1 H = high SRBC selected; L = low SRBC selected. 2 Distance in mm between sample well center and precipitate meniscus. 3 Gen 8, Gen 19 = generations 8 and 19. 4 IgM was undetectable in 3 samples. FIGURE 3. Individual SRBC agglutinin titers, sex means, and their connection for control, high SRBC selected, and low SRBC selected lines. Data for the control line were available at generations 8 and 19 only. intermediate, and line L was lowest for each agglutinin type. Gender had an important effect on SRBC titer. Over all generations, mean titers of hens were higher (6.6, n = 127) than those of cocks (5.1, n =113; F = 4.87, P = 0.028). Individual values are shown in Figure 3, which also includes the location of the gender means and a connecting line. RRBC titers were higher in hens compared with cocks (dilution corrected, 5.6 vs. 5.4), but this difference was not significant, F = 2.4, P = 0.13. Qualitative differences in the forms of the agglutination of rabbit cells accompanied the observed quantitative changes. These were easily recognized, and a scoring system was used to categorize settling patterns. The 3 distinctly positive agglutination types were found reinforce the point that these antibodies represent a heterogeneous population. Type 4 reactions resembled buttons (which typically represent an absence of agglutination in the SRBC system) and were very weak and suspended easily after abrupt manual agitation of the microtiter plates. For the purpose of calculating line means, these values were regarded as negative and recorded as zero. Thus lower scores (except zero) result from higher antibody avidity. A significant difference was found with line H having a lower score than the control line or line L (Table 1). Qualitative differences in IgM were detected among lines using Ouchterlony double diffusion data (Table 2). TABLE 1. ANOVA of RRBC agglutination type line mean scores Line1 H Control L n 100 40 100 Mean agglutination score2 x 1.90 2.25x 2.41y SD 0.72 0.78 0.85 Means sharing no common superscript differ significantly (P < 0.05). H = high SRBC selected; L = low SRBC selected. 2 Where 1 = classic, blanket pattern, 2 = granular or grit, 3 = annular or ring, 4 = very weak reactions, cells easily suspended. x,y 1 Measurements of the locations of the precipitates indicated that the molecular weight of IgM in lines H and L was slightly higher than in the control line, which was because their precipitates were located closer to the sample wells than those of the control line, indicating a slower diffusion rate. The average precipitate location of the control line at generation 8 was 6.9 mm compared with 6.1 and 6.0 for lines H and L, respectively. The corresponding measurements for generation 19 were 5.9 mm for the control and 5.4 mm for lines H and L. The apparent lower diffusion distances found in generation19 could not be a consequence of SRBC selection because they were present in the control line as well. It is more likely that the observed generational difference was an artifact related to slight variations associated with the assay method. However it is important to note that the generational differences were consistent in that the control line had higher values at both times. DISCUSSION In humans, some of the mechanisms of innate immunity are based on cross-reacting natural antibodies. Examples include the hyperacute rejection of xenografts and complement-dependent cell destruction. The best-characterized target antigen in xenograft rejection is α (1–3) galactose, which is also widely distributed on the surfaces of bacteria, protozoa, and viruses (Starzl and Zinkernagel, 1999). Apart from carbohydrates, natural antibodies directed to a variety of exogenous and endogenous antigens, including proteins, have been described (Avrameas, 1991). Natural antibodies that agglutinate rabbit cells are likely to do so because these cells express high levels of Galα1-3Galβ1-4GlcNAc-R, known as the α-Gal epitope (Galili, et al., 1984). They are widely distributed among animals and are found in several species of fowl including chickens, ducks, emus, and turkeys (Cotter, 1998b). In primates such antibodies are known as anti-Gal, and presumably those described here represent the chicken equivalent. Their source is not known for certain, but natural antibodies have long been believed to originate from a stimulus provided by the surface of gut microor- 224 COTTER ET AL. ganisms (Weiner, 1951). In human serum anti-Gal is found between 30 and 100 µg/mL representing 1% of all circulating antibody, and it is the largest single kind. It is likely to be as important in chickens but is presently incompletely characterized. However the following properties have been observed. The agglutination of rabbit cells of inbred Leghorn chickens showed galactose concentration-dependent inhibition. Increasing galactose in the diluent (from 0 to 1% in 0.25% intervals) was accompanied by a highly significant reduction in hemagglutination score. In a limited study of pooled serum from the lines described here RRBC agglutination was reduced in the presence of galactose, melibiose, and cellobiose. The levels of RRBC agglutinins of turkey sera were compared by using 2 diluents (PBS vs. 1% galactose in PBS). The regression of hemagglutination score in the presence of galactose vs. its absence was negative and highly significant (P. Cotter, unpublished data). Collectively these observations support the idea that these antibodies are related to the anti-Gal of primates. Anti-Gal isotypes of the IgG, IgM, and IgA classes are found in primate serum, and IgG is the predominant type (Galili, et al. 1995). Because chickens have isotype classes corresponding to those found in mammalian species (Benedict, and Berestecky, 1987), the rabbit agglutinins described here are likely to have broad class membership as well. Some rabbit agglutinins found in chickens were inactivated by 2-mercaptoethanol (Cotter, 1998a), suggesting that that IgM isotypes are included, although some chickens also have 2-mercaptoethanol–sensitive IgG (Delhanty, and Solomon, 1966). Hemagglutination depends on many variables, including the amount and type of antibody present; the size, number, and location of available antigen sites; and the pH, temperature, and ionic strength of the test medium (Quinley, 1993). Those antibodies described here represent a heterogeneous system because some rabbit agglutinins are removed by guinea pig cells, whereas others are not (Bailey, 1923). Because a single diluent (PBS) and temperature were used throughout, conditions governing agglutination were probably optimal for only a portion of the antibodies present in any one line at any generation, which would account for some of the difference in the observed patterns. Some of the genes that control the SRBC response are located on the sex chromosome (Boa-Amponsem, et al, 1997), which may account for the gender differences observed here. Although the gender difference in rabbit agglutinin levels was not significant females had higher titers than males. Thus the RRBC observations paralleled the SRBC results. Perhaps there are general immunoregulatory genes located on the Z chromosome that influence a broad range of antibody specificities. If such genes exist they must act to down regulate antibody production because the hemizygous (Zw) females have higher titers. Based on the observations of the control line it is likely that any major gender difference in genes controlling rabbit agglutinin levels were long ago removed by natural selection. This observation may reflect the relative impor- tance of the anti-Gal antibody in the overall immunity scheme, which is why natural sheep agglutinins are at low levels in unsensitized chickens, but rabbit agglutinins are at moderate levels (Bailey, 1923). Recently, microsatellite regions for specific antibody responses to SRBC were detected on chromosomes 3, 5, 16, and 23 by half sib analysis, and 10, 16, and 27 by line cross analysis (Siwek et al., 2003). Whether these regions are also involved in anti-Gal or antibodies to lipopolysaccharides, and protein antigens is being examined. A comparison of the responses of the control line suggests that rabbit agglutinin levels are more sensitive to environmental influences compared with sheep agglutinins. However, to our knowledge there was no dramatic change with respect to husbandry or feed between these generations. There was virtually no change in SRBC titers of the control line between generations 8 and 19 but its rabbit titer was near that found in line H. The possibility that the present results can be accounted for because of epitopes common to both erythrocytes is unlikely. RRBC antibody titers declined while SRBC titers rose through generation 14, and in the original survey data (Bailey, 1923) low levels of SRBC agglutinins were found in the presence of much higher RRBC levels. Both observations would be difficult to explain if the 2 erythrocyte species shared epitopes. It would be presumed that the heritability of the antiGal response would be lower than the SRBC response. Thus artificial selection for anti-Gal titers would likely be met by more resistance form natural selection and progress in either direction would be slower than observed for sheep antibody. Our data indicate that directional selection for SRBC antibody was highly successful and it was accompanied by an indirect effect on natural rabbit agglutinins. When all data are included the correlation between SRBC antibody titer and RRBC titer was 0.43 (P = 0.0). This must be considered a moderately high value given that the indirect response was biphasic and partly opposite in direction to that of the direct response. In addition the SRBC levels of the control line remained virtually unchanged over the course of the experiment but RRBC levels increased. Natural antibodies may play an important regulatory (often enhancing) role in specific immune responses (Ochsenbein et al., 1999). The high levels of natural antibodies to proteins in the present lines (Parmentier et al., 2004) and the functional relation between natural and specific antibodies (Lammers et al., 2004) are in agreement with this idea. It is difficult to speculate about the biological significance of anti-Gal antibodies in poultry, although a role in anti-retroviral resistance has been proposed in primates. It is noteworthy that levels of natural antibodies binding proteins are very low in young chickens, it remains to be established how age affects anti-Gal antibodies. In summary the selection results show clear evidence for the integration of genes that control the natural and acquired immune systems. The exact nature of equiva- ANTIBODY SELECTION AFFECTS NATURAL RABBIT AGGLUTININS lence between the natural rabbit antibody described here and the anti-Gal of primates awaits further experimentation. Given the established importance of anti-Gal in primates it is possible that the anti-rabbit agglutinins described here represent a significant component of the immune system of chickens. Further study of the properties of chicken anti-Gal is clearly indicated. REFERENCES Avrameas, S. 1991. Natural antibodies: from horror autotoxicus to gnothi seauton. Immunol. Today 12:154-159. Bailey, C. 1923. A study of the normal and immune hemmagglutinins of the domestic fowl with respect to their origin, specificity and identity. Am. J. Hyg. 3:370–393 Benedict, A. A., and J. M. Berestecky. 1987. Special features of avian immunoglobulins. Pages 113–125 in Avian Immunology Basis and Practice. Vol. I. A. Toivanen and P. Toivanen, ed. CRC Press, Boca Raton, FL. Biozzi, G., D. Mouton, O. A. Sant’Anna, H. C. Parssons, M. Gennari, M. H. Reis, V. C. A. Fereira, A. M. Heumann, Y. Boothillier, O. M. Ibanez, and A. Kleinschmidt. 1979. Genetics of immunoresponsiveness to natural antigens in the mouse. Curr. Topics Microbiol. Immunol. 85:31–98. Boa-Amponsem, K. E., A. Dunnington, and P. B. Siegel. 1997. Genetic architecture of antibody responses of chickens to sheep red blood cells. J. Anim. Breed. Genet. 114:443–449. Bordet, J., 1898. Sur l’agglutination et la dissolution des globules rouges par le serum d’animaux injectes de sang defibrine. Ann. Inst. Pasteur (Paris) 12:688–695. Bovenhuis, H., H. Bralten, M. G. B. Nieuwland, H. Parmentier. 2002. Genetic parameters for antibody response of chickens to sheep red blood cells based on a selection experiment. Poult. Sci.. 81:309–315. Cotter, P. F. 1998a. Denaturation kinetics of chicken hemagglutinins using 2-mercaptoethanol and heat. Poult. Sci. 77(Suppl. 1):100. (Abstr.) Cotter, P. F. 1998b. Naturally occurring rabbit erythrocyte agglutinins in fowl sera. Poult. Sci. 77(Suppl. 1):100. (Abstr.) Cotter, P. F. 2000. Quantitative and qualitative heterogeneity of bile IgA in commercial laying hens. Pages 78–82 in Proceedings of the 6th Avian Immunology Research Group Meeting. College of Veterinary Medicine, Cornell University, Ithaca, NY. Crowle, A. J. 1973. Pages 1–64 and 247–303 in Immunodiffusion 2nd ed. Academic Press, New York. Delhanty, J. J. and J. B. Solomon. 1966. The nature of antibodies to goat erythrocytes in the developing chicken. Immunology 11:103–113. Falconer, D. S. 1981. Introduction to Quantitative Genetics. 2nd ed. Longman Group Ltd., Harlow, Essex, UK. 225 Galili, U., O. Minanov, R. E. Michler, and K. R. Stone. 1995. High affinity anti-Gal IgG in chronic rejection of xenografts. Xenotransplantation 4:127–131. Galili, U., E. A. Rachmilewitz, A. Peleg, and I. Flechner. 1984. A unique natural human IgG antibody with α-galactosyl activity. J. Exp. Med. 160:1519–1531. Lammers, A., M. E. V. Klomp, M. G. B. Nieuwland, H. F. J. Savelkoul, and H. K. Parmentier. 2004. Adoptive transfer of natural antibodies to non-immunized chickens affects subsequent antigen-specific humoral and cellular immune responses. Dev. Comp. Immunol. 28:51–60 Ochsenbein, A. F., T. Feher, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, R. M. Zinkernagel. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science 286:156–159 Parmentier H. K., A. Lammers, J. J. Hoekman, G. De Vries Reilingh, I. T. Zaanen, H.F. Savelkoul. 2004. Different levels of natural antibodies in chickens divergently selected for specific antibody responses. Dev. Comp. Immunol. 28:39–49 Pinard, M.-H., A. M. van Arendonk, M. G. B. Nieuwland, and A. J. van der Zijpp. 1992. Divergent selection for immune responsiveness in chickens. Estimation of realized heritability with an animal model. J. Anim. Sci. 70:2986–2993. Pinard-van der Laan, M.-H., P. B. Siegel, and S. J. Lamont. 1998. Lessons from selection experiments on immune response in the chicken. Poult. Avian Biol. Rev. 9:125–141. Pouthoulakis, C. 1999. Enterotoxin A of Clostridium difficile and α-Gal epitopes. Pages 215–228 in α-Gal and Anti-Gal, Subcellular Biochemistry volume 32. U. Galili and J. L. Avila, ed. Kluwer Academic/Plenum, New York. Quinley, E. D. 1993. Immunohematology Principles and Practice. J. B. Lippincott Co., Philadelphia Siwek, M., S. J. B. Cornelissen, M. G. B. Nieuwland, A. J. Buitenhuis, H. Bovenhuis, R. P. M. A. Crooijmans, M. A. M Groenen,G. de Vries Reilingh, H. K. Parmentier, J. J. van der Poel. 2003. Detection of QTL for immune response to sheep red blood cells in laying hens. Anim. Genet. 34:422–428. Starzl, T. E., and R. M. Zinkernagel.1998. Antigen localization and migration in immunity and tolerance. N. Engl. J. Med: 339:1905–1913 Van der Zijpp, A. J., and Nieuwland, M.G. B. 1986. Immunologic characterization of lines selected for high and low antibody production. Pages 211–215 in 7th European Poultry Conference. Vol. I., Paris. Weiner, A. S. 1951. Origin of naturally occurring hemagglutinins and hemolysins. A review. J. Immunol. 66:287–295. Yunis, R., A. Ben-David, E. D. Heller, and A. Cahaner. 2002. Genetic and phenotypic correlations between antibody responses to Escherichia coli, Infectious Bursa Disease virus (IBDV), and Newcastle Disease virus (NDV), in broiler lines selected on antibody response to Escherichia coli. Poult. Sci. 81:302–308.