versão On-line ISSN 1806-9061
Rev. Bras. Cienc. Avic. v.5 n.2 Campinas maio/ago. 2003
Schmidt GSI,II; Hellmeister Filho PIII; Zanella ELIV
IAnimal Scientist, PhD, Embrapa Suinos e Aves
IIIDVM, MSc, Fapesp scholarship
IVDVM, PhD, Embrapa Suínos e Aves and FAMV/UPF, RS
The objective of this study was to evaluate the effect of selection for body weight on the genetic variability and diversity in broiler lines. Two paternal broiler lines (LL and LLc) were used. LL line was selected for 12 generations for growth and carcass and reproduction characteristics. The LLc line was established from LL line in 1985 and mated at random. Blood samples from six chickens per line were collected and used for molecular analysis. Also, a DNA pool was made for each line to compare effects between lines. Data were analyzed considering the collected information on the presence or absence of DNA bands. Band sharing scores were calculated using the DICE coefficient. The pattern of the 21 most representative bands was used. DNA fingerprinting (DFP) showed 90.48 % of polymorphism bands for both lines. Difference between lines was not due to the presence or absence of bands, but to the frequency of such bands in each genotype. Considering that both lines had the same genetic background, changes on band frequency were probably due to selection. Selection for body weight had an effect on the band frequency as evaluated by DFP, and for this reason this technique could be used as a tool in the selection process. Results also suggest that bands 4, 5 and 19 were linked to body weight traits, and bands 9, 10, 12, 13 and 21 were linked to reproductive traits such as egg production.
Keywords: body weight, fingerprinting, molecular markers, variability.
A great genetic diversity exists on germoplasms and, therefore, there is considerable potential to improve economically important traits of animals. Nevertheless, breeding programs are limited in that there is a lack of definite markers associated to relevant traits that could help in the selection process. Genotype superiority might be efficiently and quickly identified by the use of molecular markers and, consequently, extra gains could be obtained as well as reduction in cost, time, energy and space.
The genetic gain for the main traits of economic interest might be maximized if known quantitative genetic techniques are associated to the use of genetic markers that enable the identification of animals with higher productive potential from a genetic point of view. Monogenic markers have been used successfully in the selection of traits that are genetically more complex or difficult to evaluate, among which: morphological traits of monogenic inheritance, proteins (isoenzymes) and, more recently, DNA polymorphisms (Tanksley et al., 1981; Guse et al., 1988).
Efforts have been directed to obtain molecular markers in order to map and evaluate genetic variability and diversity in many species. Thus, the first studies were based in restriction fragment length polymorphisms (RFLP), random amplified polymorphic DNA (RAPD) and single sequence repeats (SSR).
The analysis of DNA polymorphisms using restriction fragments or other techniques is a tool to evaluate genetic diversity, germoplasm characterization and to construct genetic maps. Such information establishes many levels of selection possibilities. The DNA fingerprinting technique (DFP) developed by Jeffreys et al. (1985) is effective to identify individuals in the population and to evaluate the genetic relationship between and within populations (Burke & Bruford, 1987).
DFP using multilocus probes may generate genomic information and, thus, permit the characterization and identification of genotypes. In birds, the technique was first used in wild birds for species characterization of individuals (Wetton et al., 1987).
Many studies were performed with DFP to characterize genotypes. Dunnington et al. (1990) evaluated the differences between strains divergently selected for growth rate and reported specific bands for the populations with high and low growth rates, indicating that some bands were related to this trait.
As a measurement of the genetic distance between strains, DFP may be used to estimate heterosis and to direct mates (Haberfeld et al., 1996). Gavora et al. (1996) reported high correlations (0.68 to 0.87) between the ratio of band sharing (BS) and heterosis for sexual maturity, egg production and adult body weight.
Regression analysis between pedigree and BS showed a correlation of 0.992 (Zhu et al., 1996), and thus, is a good indicator of inbreeding. Thus, if pedigree is not available, DFP might help to monitor the inbreeding index of a population, as well as to direct mates towards reduction of the detrimental effects that result from inbreeding (Zhu et al., 1996).
This study used DFP to evaluate the effects of selection for body weight on the variability and genetic diversity of broiler strains.
MATERIAL AND METHODS
Two experimental male broiler strains were used. LL and LLc strains were kept in the bird genetic selection program from Embrapa Suínos e Aves. Strain LL was selected for 12 generations for body weight and feed conversion, using mass selection, and for fertility and hatchability, using independent culling levels (male and female breeders with the poorest performance were culled). More emphasis was given to body weight, which represented 80% of the selection intensity. A mean of 30 sire families were used in each generation, mated to 8 females each. Pedigree control was used to prevent an increase in inbreeding index. Body weight selection was performed at 35 days of age and pre-selected birds were evaluated for feed conversion from 35 to 42 days. Mean selection intensity in each generation was 2.0% for males and 16% for females.
Control strain LLc derived from LL in 1985 and was kept at random mating. Therefore, both strains originated from the same genetic background and any differences were in consequence of the selection process that was used.
DNA Extraction and DFP preparation
Blood samples (1 mL) were taken from the brachial vein of 6 birds/line using disposable syringes with 2% EDTA as anticoagulant (0.05 mL/mL of blood). Samples were transferred to sterile microcentrifuge tubes and stored at -20 ºC.
DNA samples from both lines were analyzed. Also, samples from each line were pooled for analysis between lines. DNA was extracted from the erythrocytes by osmotic shock, successive deproteinizations and, finally, precipitation and washed with ethanol. It was added 100 mL of SSC 1x (NaCl 3M; sodium citrate 0.3M) in 20 mL of blood, mixed and centrifuged for 2 min at 15,000 rpm. The supernatant was carefully removed, cells were resuspended in 100 mL SSC 1X and the procedure was repeated. The supernatant was discarded and 500 mL sodium acetate (0.2M pH 5.5) was added and accumulations were reduced as much as possible.
Approximately 50 mL of 10% SDS (sodium dodecyl sulphate) was added to a final concentration of 1% and the solution was mixed by inversion. Phenol:chloroform (1:1) was added (500 mL) and the aqueous phase was mixed with the organic phase until complete homogenization. The tube was centrifuged for 5 min at 15,000 rpm, and the supernatant was carefully transferred to a fresh microcentrifuge tube. The same volume of phenol:chloroform was added again and the steps were repeated until a clear supernatant was obtained.
DNA was precipitated with 1 mL of ethanol 100%. After mixing by inversion, it was centrifuged for 2 min at 15,000 rpm and the ethanol was discarded. The precipitate was washed with 500 mL ethanol 70%. After ethanol was discarded, the precipitate was dried at 37 ºC and then resuspended in 150 mL TE (10mM Tris-HCl pH8.0; 1mM EDTA pH 8.0).
Autoradiograms for each of the DNA samples were visually analyzed for the presence and absence of bands.
Similarity within and between lines was determined based on binary data. Band sharing (BS) or similarity indexes were calculated based on the DICE coefficient, which represents the proportion of shared bands to total band number in the profiles when comparing the lines, i.e.:
BS = Nab/((Na+Nb)/2) or 2xNab/[(2xNab)+Na+Nb]
Where: BS bandsharing index between two samples; Nab number of bands shared by samples; Na number of bands present on sample A; and Nb number of bands present on sample B.
Calculated BS was used to produce the dendogram with the Numerical Taxonomy and Multivariate Analysis System (Rohlf, 1992).
RESULTS AND DISCUSSION
Table 1 presents the results of DFP band analysis based on autoradiogram profiles, the occurrence frequency (f) of each band within the line and the deviation between the selected line and control line. Complex polymorphic profiles were seen for 41 bands. Band profiles were different between samples and the 21 most representative bands were selected for profile evaluation.
Except for bands 2, 9 (LL), 3 and 11 (LLc), all bands were present in the DNA, and DFP showed 90.48% of polymorphic bands for both lines. Therefore, differences between lines are not due to the absence or presence of bands, but due to the frequency of such bands in each genotype.
Frequency distribution in LL was higher for bands 1, 5, 8, 11, 12, 14 and 16 (50%), 3, 6, 7, 10, 13 and 20 (33.3%), whereas distribution in LLc was between 66.7% (4, 6, 7, 8, 10, 12, 13, 16 and 20) and 50% (1, 2, 7, 9, 14, 15, 17 and 18).
The lines had the same genetic background; therefore, the differences seen in DFP should be credited, for the most part, to the effects caused by selection. Considering body weight at 42 days of age, the deviation between line LL (2,375g) and LLc (1,810g) was as much as 565g. This represents the genetic gain that was obtained after 12 selection generations, i.e., approximately 47g/generation. Inbreeding could be considered a factor causing changes, but it was controlled by directed mating in line LL and inbreeding was around 0.17% in each generation, which is a low value to cause marked differences.
Bands 4 (33%), 5 (33%) and 19 (50%) showed the greatest change in frequency. Therefore, these bands must be linked to body weight. Apparently, selection acts fixing bands and reducing variability. This process might have occurred in band 4, which was present in 100% of the samples. Besides, if band 4 were linked to growth traits, it would no longer be important from the selection point of view, since it had already been fixed in the line. Considering the existence of negative genetic correlation between body weight and reproductive traits (egg production, fertility and hatchability), it should be expected a reduction in the frequency of some bands as a consequence of the selection process, as seen for bands 9, 10, 13 and 21. In the cases that small or absent change in frequency occurred, such as in bands 1, 8, 12, 14, 15 and 16, it might be inferred that these bands were linked to traits non-correlated to growth traits.
The similarity matrix was calculated based on the DICE index and on the similarity values within the line (Table 2). Similarity values (BS) within line LL varied from 15.4% (samples 1 and 4) to 66.7% (samples 2 and 4), with a mean value of 40.13%. Within line LLc, BS varied from 34.8% (samples 9 and 11) to 72.0% (samples 12 and 13), with a mean value of 52.37%. If BS were considered an indicator of genetic variability, these results would suggest that selection per se could not be responsible for the reduction in genetic diversity. This would probably explain why the genetic gain for the economically interesting traits was maintained throughout selection, corroborating results reported by Dunnington et al. (1994). On the other hand, the results might indicate that DFP would not be effective for estimating genetic diversity.
Body weight at 42 days and age at sexual maturity, which are traits with high to moderate heritability, are generally superior when the progeny was originated from parents with lower BS (Haberfeld et al., 1996). On the other hand, the results obtained here showed that band frequency changed as a function of the selection process, permitting to identify the relationship among band frequency and traits of economic interest. Thus, it is suggested that the evaluation of the genomic similarity between individuals based on information provided by molecular markers may have an important role in animal breeding programs, such as recurrent selection or even in a screening process of individuals for population growth. Besides, this methodology might be used for sex-linked traits (egg production), for traits that are evaluated after animal slaughter (carcass) and for traits that are expressed late in the animal development (age at sexual maturity, egg production).
The commercial production of broilers and eggs uses complementarity and heterosis effects resulting from different types of mating. Currently, there is no available method to predict the level of complementarity and heterosis that result from mating besides progeny evaluation, which is a costly and time consuming process. Gavora et al. (1996) reported high correlations (0.68 to 0.87) between BS ratio and heterosis for traits such as age at sexual maturity and adult body weight. Besides, mates between individuals with higher BS differential would be the source of variability within a population.
Inbreeding is another important parameter in breeding programs. The inbreeding coefficient is the probability of two genes at a locus being identical as a result of descent. It is estimated based on the pedigree or on the size of population and the variation on the size of male and female families used for mating (Falconer, 1982). Nevertheless, such estimate may not reflect the true inbreeding of populations, since it is affected by other factors. Thus, complete records of the population are required for determining inbreeding and these might not be available. Many studies have shown that similarity determined by DFP within a line is a good indicator of inbreeding. Kuhnlein et al. (1990) reported that band frequency and allelic frequency estimated by DFP were highly correlated to the inbreeding coefficient in birds, which was confirmed by the correlation of 0.992 between the inbreeding coefficient and the BS obtained by Zhu et al. (1996). Therefore, BS might substitute inbreeding coefficient to orient mating. This could reduce the detrimental effects that mating orientation might have on economic traits, such as egg production, fertility and hatchability, and on selection indexes based on relationship matrices (animal model).
Selection for body weight affected the frequency of bands determined by DFP and this technique can be used as a tool in the selection process. Bands 4, 5 and 19 are possibly related to body weight.
Burker T, Bruford MW. DNA fingerprinting in birds. Nature 1987; 327(6118):149-152. [ Links ]
Dunnington EA, Gal O, Plotsky Y, Haberfeld A, Kirk T, Goldberg A, Lavi U, Cahaner A, Siegel PB, Hillel J. DNA fingerprinting of chickens selected for high and low body weight for 31 generations. Animal Genetics 1990; 21:246-257. [ Links ]
Falconer DS. Introduction to quantitative genetics. Burnt Mill, Longman House; 1982. [ Links ]
Gavora JS, Fairfull RW, Benkel BF, Cantwell WJ, Chambers JR. Prediction of heterosis from DNA fingerprinting in chickens. Genetics 1996; 144(2):777-784. [ Links ]
Guse RA, Coors JG, Drolsom PM, Tracy WF. Isozyme marker loci associated with cold tolerance and maturity in maize. Theoretical and Applied Genetics 1988; 76:398-404. [ Links ]
Haberfeld A, Dunnington E, Siegel PB, Hillel J. Heterosis and DNA fingerprinting in chickens. Poultry Science 1996; 75(8):951-953. [ Links ]
Jeffreys AJ, Brookfield JFY, Semenoff F. Positive identification of an immigration test-case using human DNA fingerprinting. Nature 1985; 317(6040):818-819. [ Links ]
Kuhnlein U, Zadworny D, Dawe Y, Fairfull RW, Gavora JS. Assessment of inbreeding by DNA fingerprinting: development of a calibration curve using defined strains of chickens. Genetics 1990; 125:161-165. [ Links ]
Rohlf FJ. Numerical taxonomy and multivariate analysis system - NTSYS -pc. New York: Exeter Software, 1v, 1992. [ Links ]
Tanksley SD, Medina Filho M, Monseieur R, Brocas H, Lequare AS, Christophe DA. The effect of selection on metric characteres in a specific backcross of tomato - basis of an early screening procedure. Theoretical and Applied Genetics 1981; 60:291-296. [ Links ]
Wetton JH, Carter RE, Parkin DT, Walters D. Demographic study of a wild house sparrow population by DNA fingerprinting. Nature 1987; 326(6118):147-149. [ Links ]
Zhu J, Nestor KE, Moritsu Y. Relationship between band sharing level of DNA fingerprinting and inbreeding coefficients and estimation of true inbreeding in turkey levels. Poultry Science 1996; 75:25-28. [ Links ]
Gilberto Silber Schmidt
Embrapa Suínos e Aves
BR 153 Km 110, Caixa Postal 21
Phone: +55 +49 442-8555
Fax: +55 +49 442-8559
Arrived: june 2002
Approved: january 2003