In the present study we analyzed a large number of healthy goats from different holdings, in order to characterize the PRNP variability in the Greek goat population. As expected, the caprine PRNP gene displayed a high degree of variability. The non-synonymous polymorphisms found included twelve previously reported changes [3, 15, 16] and three that were novel for the Greek goat population. Of the newly detected alleles, the N146D polymorphism, which has also been found in Cypriot goats , is strongly associated with resistance against scrapie. The I218L dimorphism has been reported in both Chinese  and British goats . Finally, the W102G dimorphism was originally observed in tandem with a reduced number of octapeptide repeats (three instead of five) . In our study, however, this allele was found in tandem with the wild type five octarepeats, as reported by other recent studies [4, 21, 22]. Four known silent changes were detected in the tested population. One of these, the codon 179 synonymous change, is reported for the first time in Greek goats, but has previously been found in Cypriot , and Spanish goats .
In contrast to studies on goat populations from Northern Europe, but consistent with previous studies of Greek, Cypriot , and Italian  goat populations, P240 was found to be the most abundant PRNP allele. As in other countries, the Q211 and K222 alleles were found in linkage with the S240 allele, and they were also frequently linked with the minor alleles of silent changes at codons 42 (a) and 138 (c). Although silent changes do not affect the PrP amino acid composition, their importance should not be discounted, since they may be linked with RNA processing or with genetic loci associated with important production traits. Indeed, a recent publication on Chinese goats  reported the linkage of codon 42 with milk traits in some goat breeds.
Interestingly, the Q211, S146 and K222 PRNP alleles, for which a strong protective effect against goat scrapie was suggested by field [4–7] and in vitro studies , and further confirmed by recently published studies on experimentally challenged goats [8–10] and on transgenic mice , were detected in relatively high frequencies in the tested population. This result depicts the special situation of Greek goats with respect to the frequency of the before mentioned alleles, as they represent the only goat population among European countries where all three most scrapie-protective alleles are detected. Indeed, even though the Q211 and K222 alleles have been found in relatively high frequencies in some countries [6, 24], the S146 allele was absent from these, and other populations , while in other cases the K22 2 allele was represented in very low frequencies [23, 19].
Even though goat scrapie prevalence is high in Greece, a clear association between specific caprine PRNP alleles and scrapie protection has not been determined, due to the lack of large-scale case-control studies. Nevertheless, a meta-analysis performed on the published data referring to PRNP genotypes from scrapie-affected Greek goats may lead to quite strong estimations. The analysis included 146 scrapie-positive cases from 36 herds [3, 15, 16], where the following allelic frequencies were detected: S49 (0.3%), P110 (1%), S127 (0.7%), T136 (0.3%), M142 (0.7%), R143 (0.7%), H151 (0.3%), H154 (2.4%), Q154 (0.3%), Q168 (0.7%), and K222 (1.7%)[3, 15, 16]. Importantly, no goat scrapie-cases have been reported for Q211 or S146 carriers, suggesting that the probabilities of disease occurrence in these animals are low.
The detection of K222 carriers within scrapie affected animals, which has also been reported in French goats , verifies that this allele does not confer complete resistance to goat-scrapie. It should, however, be emphasized that K222 was found in five heterozygous scrapie-positive animals, originating from three heavily affected herds, associated with a scrapie-isolate displaying a distinct molecular profile as compared to other Greek isolates . The same allele was detected in a much higher frequency (4.2%) in healthy animals (1222 animals, corresponding to relevant data from all three previously published studies and the study presented here). Even though these data do not correspond to a case-study, they still provide valuable information on the protective effect of the K222 allele. This claim is also strengthened by the fact that no scrapie cases were detected in K222 homozygous goats, even though the latter observation is only an indication that cannot be verified due to the low number of K222 homozygous animals. Taken together, the above data suggest that (a) the K222 allele displays a scrapie-protective effect in Greek goats, and that (b) a herd’s scrapie history, as well as the scrapie strain, are important factors that should be considered before selecting the most appropriate PRNP alleles for scrapie-resistance breeding.
Based on the previous data and on the results from experimental challenges in goats [8–10] and transgenic mice , it seems that Q211, S146 and K222 alleles are potential candidates for selective goat-scrapie resistance breeding programs in Greece, provided that herd history and scrapie strain information are also available. Such programs should be most efficient when bucks carrying the desired alleles are recruited. This is because in the Greek farming system the male to female ratio in a holding is usually 1:20. Thus, bucks have a strong impact on the determination of the overall genetic profile of the herd. Our results indicate that the Q211, S146, and K222 alleles are present in relatively high frequencies in both the general population, and in bucks specifically, suggesting that a scheme targeting the selection of bucks carrying these alleles is feasible. In contrast, other resistance-associated alleles (M142, R143 and H154) could not be targeted, due to their much lower frequencies in bucks. Furthermore, the H154 allele has been associated with increased susceptibility to atypical scrapie in both sheep and goats [26, 27]. Thus, the feasibility of incorporating M142, R143, and H154 alleles into a selective breeding scheme is low.
Breeding for resistance in goats by recruiting bucks harbouring the Q211, S146, and K222 alleles, should result in a rapid increase of their frequencies in the overall population, which would be expected to reduce the individuals’ risk for developing scrapie. The application of such a strategy would be expected to effect a quick halt on the disease spread within a scrapie-affected herd, and to reduce a herd’s overall risk of becoming scrapie-affected, in a scrapie-free herd. Indeed, available data on scrapie-affected flocks of sheep show that breeding for scrapie resistance with ARR/ARR rams results in rapid control of scrapie-outbreaks . Furthermore, it has been shown that a flock’s PRNP profile is a risk factor for scrapie occurrence . Increased frequencies of the resistance-associated ARR allele, as a result of selective breeding, have a population effect, reducing the scrapie infection risk even for animals of susceptible genotypes .
Such an aggressive scheme as the one described above for sheep, cannot be directly applied to Greek goat populations, because (a) there seems to be more than one allele associated with high -but not absolute- scrapie resistance, and (b) the frequency of bucks homozygous for either the Q211 and/or K222 allele is low, and no males homozygous for the S146 allele have been detected. Given that the observed frequencies of the Q211, S146, and K222 alleles are moderate to high, a selection scheme targeting carriers of all three of these alleles seems to be more feasible. Furthermore, this strategy conserves more of the original variability of the population. Variability is desired both with respect to protection against other scrapie strains affecting the resistance-associated alleles, as well as with respect to preservation of production traits possibly related to PRNP. Breeding for a homogenous population poses a risk for the emergence of an uncontrolled epidemic in the event that a strain targeting the previously resistant genotypes should predominate.
A selective breeding scheme aiming at the increase of the protective Q221, S146 and K222 alleles is expected to decrease the goat-scrapie incidence and to significantly contribute to the control of scrapie in small ruminants. This is because scrapie affected goats may serve as reservoirs for the disease agent, which can subsequently be transmitted to sheep. The possibility of this transmission to occur is thought to be high when taking into consideration that (a) the main farming system in Greece includes mixed sheep and goat flocks and that (b) the disease occurs more often in sheep than in goats . Thus, the reduction of scrapie incidence in goats, in combination with scrapie reduction in sheep due to the application of the corresponding sheep-scrapie eradication program, is expected to strengthen the result of the latter and to have a beneficial effect on the control of the disease in small ruminants in general. This is highly relevant not only for Greece, but also for other countries where sheep scrapie resistance programs are applied.
Further, the control of scrapie is also expected to improve animal welfare and to highly benefit stakeholders, by reducing costs due to animal loss and by increasing sheep and goat farming income and the dairy products related to the agricultural sector, which significantly contributes to the total agricultural income.
In summary, the present study is intended to extend current knowledge on PRNP genetic variability in the Greek goat population, one of the biggest in Europe, and to provide valuable information on scrapie resistance-associated allele frequencies found in male-goats. Further, based on the results presented here and on other studies on Greek goat PRNP alleles and their association with scrapie, we suggest the design of a goat-scrapie resistance program targeting the Q211, S146 and K222 alleles. The implementation of such a program is expected, to our opinion, to greatly aid the control of scrapie in small ruminants and to highly benefit the relevant stakeholders in Greece. Taking into account the special situation in Greece, where one of the largest goat populations in Europe resides and all three most protective goat-scrapie alleles are detected in relatively high frequencies in both the general population and in bucks, it follows that the implementation of the breeding program proposed here, seems feasible. Moreover, the application of the suggested scheme in Greek goats would serve as an example of the results such a strategy would have on the control of the disease in small ruminants and of its beneficial effects on relative stakeholders in general.