A New AS-PCR Method to Detect CSN201 Allele, Genotyping at Ca-Sensitive Caseins Loci and Milk Traits Association Studies in Autochthonous Lazio Goats

Simple Summary Identifying new alleles and mutations at calcium-sensitive casein loci requires continuous updating of genotyping protocols useful for association study. This study reports (1) a new, more efficient, specific PCR-based genotyping protocol to detect the CSN201 allele in goats, (2) the genetic characterization at the CSN1S1, CSN2, and CSN1S2 loci in three endangered goat breeds reared in the Lazio Region (Central Italy) and, (3) the association of the genotypes observed in the studied animals with parameters that might affect the milk’s traits. As regards the CSN1S1, CSN2, and CSN1S2 loci, no animals were found to be carriers of the CSN1S101, CSN1S1E, CSN201, CSN1S2D, and CSN1S20 alleles; instead, for the CSN1S1 locus, a high frequency of alleles associated to a low (CSN1S1F) and high (CSN1S1A*,B*) content of the αs1 casein (αs1-Cn) content in milk, with CSN1S1F≥ CSN1S1B* ≥ CSN1S1A* being observed. An association between the different genotypes at the CSN1S1 locus and some milk traits, namely the fat and protein yielded and the fat, protein, solids-not-fat, and casein percentages without an effect on the milk yield, was observed. Abstract Calcium-sensitive caseins are the main protein component of milk. In the goat, they are encoded by three genes (CSN1S1, CSN2, and CSN1S2) located on chromosome 6. A high number of alleles has been discovered for these genes in the goat species, responsible for changes in the milk’s qualitative and quantitative characteristics. This study aimed to develop an Allele-Specific PCR (AS-PCR), which allowed us to unequivocally detect goat carriers of the CSN201 allele. Subsequently, the calcium-sensitive casein loci genotype was investigated in three native goat breeds of the Lazio Region (Bianca Monticellana, Capestrina, and Ciociara Grigia). No individuals were carriers of the CSN1S101, CSN1S1E, CSN201, CSN1S2D, and CSN1S20 alleles, while a high frequency of the alleles CSN1S1F and CSN1S1A*,B* was observed. Association analyses between the different genotypes at the CSN1S1 locus and some milk traits, namely the fat and protein yielded and the fat, protein, solids-not-fat, and casein percentages without an effect on the milk yield, were observed.


Introduction
Recent achievements in molecular genetics provide the opportunity to investigate genomic regions that directly or indirectly influence animal production. Moreover, special

Farms and Animals
Five farms located in the Latina and Frosinone Provinces, close to each other, were included in this study. The animals were reared following the same traditional management practices of the area: the goats are left to graze in daylight hours (6/8 h/day) and return to the shed at sunset, leaving the dams with their kids at night. The kids are breastfed for up to 45 ± 5 days postpartum. Manual milking is carried out twice a day, in the morning and in the afternoon, starting from weaning to drying off (at about five months). Information concerning the parity number was also available. A total of 188 goats (125 Monticellana, 27 Capestrina, and 36 Ciociara Grigia) were used for the present study. All the animals were enrolled in the Official Birth Register (ASSONAPA) and minimally related. Blood samples were collected (19 males and 169 females) for genetic characterization. Individual milk samples (50 mL) from 169 goats (112 Bianca Monticellana, 27 Capestrina, and 30 Ciociara Grigia, respectively) at three lactation stages (60,90, and 120-days postpartum, from May to August 2019) were collected in the morning, to evaluate the effect of casein polymorphisms on the milk yield and quality.

DNA Extraction
DNA was extracted from the blood by use of a Wizard DNA extraction kit (Promega-Madison, WI, USA), following the manufacturer's instructions.

Genotyping at the CSN1S1 Locus
Genotyping at the CSN1S1 locus was carried out in stages with an initial XmnI PCR-RFLP [17] to check which group of alleles the tested animals were carriers of: A*-derived alleles (A, A2, A3, A , I, G, M, 02, H, and 01), B* alleles (B , B1, B2, B3, C, D, D1, L, and E), and F or N alleles. DNA from subjects found to carry the A* group allele was subsequently analyzed by AS-PCR [18] to verify if they carried the 01 allele, while DNA from subjects found to carry the B* group allele was genotyped by the technique described by Jansá Perez et al. [16] to distinguish subjects in which the E allele was present.

Genotyping at the CSN1S2 Locus
Genotyping of the D and 0 alleles at the CSN1S2 locus was performed by the NcoI PCR-RFLP method, according to [14].

Genotyping at the CSN2 Locus
In order to identify carriers of the goats' CSN2 01 allele at the DNA level, a new AS-PCR was set up. Sequence primers used for the AS-PCR are listed in Table S1. All primers were designed with DNASIS-Pro version 3.0 software (Hitachi, Tokyo, Japan) using the goats' CSN2 sequences as templates (GeneBank, nos. AJ011018, AJ011019.3). The length of the amplified fragment-spanning intron 6 (partial) to exon 7 (partial) is 463/464 bp. Amplifications were performed in a 25-µL volume containing 100 ng of genomic DNA, a 1× PCR buffer, 1.5 mM of MgCl2, 200 µM of each dNTP, 10 pmol of each primer, and 1 U of GoTaq ® G2 Flexi DNA Polymerase (Promega-Madison, Fitchburg, WI, USA).
The thermal conditions were: 97 • C for 2 min, 30 cycles at 94 • C for 30 s, annealing at 52.5 • C for 45 s, and extension at 72 • C for 1 min. A final extension was carried out at 72 • C for 10 min.
The amplification products were later verified by electrophoresis on a 2% agarose gel (Bio-Rad, Hercules, CA, USA) in a 0.5X TBE buffer and stained with SYBR ® green (Lonza Rockland, Inc., Rockland, ME, USA).

Milk Analyses
The amount of milk (milk yield) from the morning milking was recorded on the farm. The quality parameters analyzed were: percentages of fat, protein, total casein, lactose and solids-not-fat, and amount of urea and somatic cells.  [29].
Moreover, fat, protein, total caseins, lactose, and solid-not-fat yields were calculated.

Statistical Analysis
Allele frequencies were calculated by simple allele counting [30]. Possible deviations of genotypic frequencies from expectations were tested by a chi-square test to verify if the population was in the Hardy-Weinberg equilibrium. Moreover, some population genetic indices, namely gene heterozygosity (He), gene homozygosity (Ho), effective allele numbers (Ne), and the Fixation Index (FIS), were obtained by POPGENE32 software version 1.32 (PopGene: Microsoft Window-Based Freeware for Population Genetic Analysis, Edmonton, AB, Canada) [31]. The Polymorphic Information Content (PIC) was calculated according to Botstein et al. [32]. A first statistical analysis was carried out to estimate the effect of detected polymorphisms on the milk yield and composition traits of all the animals considered as a single population. In detail, a mixed model for the repeated measures [33] implemented with SAS software (SAS 9.2 Institute, Inc., Cary, NC, USA) was used to assess the possible relationship between CSN1S1 polymorphisms and performance traits under study. Milk production data were considered as repeated measures, and the correlations between the measures in the same individual were considered in the statistical model. The statistical model included the genotype as the fixed effect (six levels), days in milk as the fixed effect of the lactation stage (3 intervals of 30 days each), the fixed effect of the breed (three levels), the fixed effect of the parity (two levels, 1st-2nd and 3rd and later), the random animal effect, and the residual error term. Subsequently, the same animals were grouped based on their genotype at the CSN1S1 locus into three different clusters, called strong (animals carrying A*A*, A*B*, or B*B* genotypes), intermediate (animals carrying FA* or FB* genotypes), and weak (FF goats). The model used was the same model described above. The values were considered significant at p < 0.05 and presented as the least squares means ± standard errors in both cases. If more than two groups were compared, a Bonferroni test was used for multiple testing.

A New AS-PCR for the CSN2 01 Allele Detection
A new fast and economical method of analysis, based on AS-PCR, was set up to identify carriers of the SNP at position 373 on the seventh exon (AJ011018:g.8915C>T) that characterizes the CSN2 01 allele [10]. For this purpose, two different allele-specific reverse primers (named CSN2N and CSN201) that differ in the last nucleotide at the 3 -end (G→A) (Table S1) were designed. Thus, for the samples without the CSN2 01 allele, PCR amplification was successful only using the reverse primer with guanine at the 3 -end, whereas the CSN2 01 homozygote samples were successfully amplified only by the reverse primer with adenine at the 3 -end. The heterozygote samples were effectively amplified with both reverse primers ( Figure 1 and Figure S2). The common forward primer (named CSN2) and the allele-specific reverse primer's sequence are part of intron 6 and exon 7, respectively, and the amplified fragment length is 463/464 bp.

Genotyping
No animals were found to be carriers of the CSN1S1 01 , CSN1S1 E , CSN2 01 , CSN1S2 D , and CSN1S2 0 alleles, applying the methods already known for the CSN1S1 and CSN1S2 genes and the new AS-PCR protocol for CSN2. Table 1 shows the allele and genotype frequencies at the CSN1S1 locus.
Among the six allele classes investigated at the CSN1S1 locus, four of them were found in the studied population: A*, B*, F, and N. CSN1S1 F (0.44) was the most common, and it was followed by alleles from group B* (0.30) and those from group A* (0.26). The CSN1S1 N allele was present in the heterozygous state in only one subject of the Bianca Monticellana breed. This finding suggests that it is almost absent in Latium goats and, therefore, cannot be considered typical of these breeds. The application of the CSN1S1 allele discrimination for the population variability evaluation is informative, being PIC, and calculated for the four alleles found: 0.58 in Bianca Monticellana, 0.51 in Capestrina, 0.58 in Ciociara Grigia, and 0.58 in overall the population.

Genotyping
No animals were found to be carriers of the CSN1S1 01 , CSN1S1 E , CSN2 01 , CSN1S2 D , and CSN1S2 0 alleles, applying the methods already known for the CSN1S1 and CSN1S2 genes and the new AS-PCR protocol for CSN2. Table 1 shows the allele and genotype frequencies at the CSN1S1 locus.

Allele Effect on Milk Parameters
The data reported in Table 2 show the effects of the six different CSN1S1 genotypes found in the studied population. No differences in terms of milk yield were observed. Significant differences among the genotypes were found in the fat and protein yielded and in the fat, protein, solids-not-fat, and casein contents. In detail, animals carrying the A*A* genotype produced more fat per milking if compared to the A*B* and B*F ones (p < 0.05). Similar results were found for the protein yield, with A*A* individuals being more productive than goats with the B*F genotype (p < 0.05). Moreover, A*A* and B*B* individuals produced milk with a greater fat content than their A*B* and FF counterparts. As per the protein, casein, and solids-not-fat percentage, the data show a statistically significant difference among the genotypes, with milk produced by the FF goat having a lower content of these constituents if compared with all the other genotypes.   Table 3 shows the effects of different CSN1S1 genotypes clustered according to the αs1-Cn content on milk yield and composition. Significant differences were observed in the fat, protein, solids-not-fat, and casein percentages. The animals carrying strong genotypes produced milk with a greater percentage of fat compared to those in the weak group (p < 0.01). Moreover, the percentage of the protein and casein were significantly higher in milk produced by individuals with strong and intermediate genotypes if compared with the weak ones (p < 0.01), with the animals belonging to the strong group showing a higher value when compared with the intermediate genotypes (p < 0.05). Consequently, the milk produced by the strong and intermediate genotypes was richer in solids-not-fat than that produced by a goat carrying a weak genotype (p < 0.01).

AS-PCR Protocol for CSN2 01 Allele Detection
From the first goat milk protein polymorphisms described by Boulanger et al. [34] for αs1 and αs2-Cn and by Dall'Olio et al. [35] for β-Cn through the application of the electrophoretic technique, an increasing number of protein variants have been discovered over the years. The development of molecular genetics technologies in recent decades has also made it possible to obtain an extraordinary amount of information about the animal genome enabling the identification of causative events of the observed phenotypic differences and the identification of new alleles.
The later application of genotyping methods (PCR-RFLP, AS-PCR, ACRS-PCR, SSCP, DGGE . . . ) has allowed a rapid and economical genotyping of individuals, regardless of the phenotypic expression, sex, and age. Identifying new alleles and mutations requires a review of existing genotyping protocols to optimize them, considering the latest findings.
From this perspective, it was necessary to set up a new AS-PCR reaction to detect carriers of the CSN2 01 allele in the goat. In fact, it has been observed that both allele-specific forward primers (5 -CGTGCTGTCCCTTTMTC -3 and 5 -CGTGCTGTCCCTTTMTT -3 ) proposed by Ramunno et al. [36] include, in the third-to-last nucleotide at the 3 -end, the transversion AJ011018.3:g.8913C>A responsible for the amino acid exchange, p.Ser166>Tyr, in the mature protein encoded by the most recently identified allele, CSN2 E ( Figure S2). This condition could reduce the efficiency and specificity of the reaction by making the genotyping method proposed by Ramunno et al. [36] ambiguous in the presence of CSN2 E variant carriers. Hence, using the method developed in this study, it is now possible to quickly genotype goats at the CSN2 locus precisely and unequivocally.

Milk Traits Phenotyping
The qualitative parameters of the milk of Lazio goats, as regards the percentages of proteins, fat, lactose, not-fat-solids, and somatic cells, are perfectly aligned with those of other native and highly selected Alpine breeds [37,38]. However, considering the differences in terms of the milk yield between native and cosmopolitan breeds, the total productions per milking (the fat, protein, lactose, caseins, and solids-not-fat yields), these goats are certainly more similar to the first ones [39,40]. These findings show that the breeding of Lazio goats has its validity in consideration of the low breeding costs and that they allow the recovery of areas where other types of livestock activities would not be economically sustainable.

Calcium-Sensitive Caseins Loci Genotyping and Population Genetic Structure
To assess the possible correlations between the CSN1S2 genotype and the milk parameters in the investigated goat populations, a genotyping was carried out to detect the null (0) and intermediate (D) alleles at this locus. No carriers of both the 0 and D alleles were found. This result was expected, as both of these alleles are rare and detected only in a few goat breeds. In fact, from the discovery of the 0 allele in the Napoletana goat breed [14], it was mainly observed in Italian goat breeds, such as Argentata dell'Etna [41], Maltese, Jonica [42], and Sarda [43], as well as in Saanen reared in the Bursa.Province in the Marmara Region of Turkey [44] and in some local Hungarian breeds [45]. Only in these latter populations, exceptionally, the CSN1S2 0 allele was observed with a relatively high incidence (0.146). Even rarer is the CSN1S2 D allele, which is still identified only in Napoletana goats (0.019) [14] and Hungarian breeds (0.005) [45].
Similarly to what was observed at the CSN1S2 locus, the analysis of the CSN2 locus showed the absence of the null allele, CSN2 01 , in the investigated populations. This result is compatible with the ones obtained in other breeds reared in Italy and characterized by the lack or very low frequencies of this null allele [10,24,36,42,43,46,47].
In this study, no analyses have been performed to detect the null allele, CSN2 0 , that, similarly to the 01 allele, is characterized by a premature stop codon (codon 58) due to a single nucleotide deletion (adenine) in a row of four adenines between nt 16 and 19 of exon 7. From its identification and characterization by Persuy et al. [48] in the Pyrenean goat breed, the presence of this allele was, in fact, no longer reported in any other goat breed.
Molecular analyses showed a fair genetic variability at the CSN1S1 locus in the goat populations studied. Four alleles (αs1-Cn A*, B*, F, and N) and seven of the sixteen possible genotypes (Table 1) were found. In particular, this study showed the absence in these populations of alleles associated with an intermediate or absent content of this protein in the milk, with only one exception for one Bianca Monticellana goat, where we found a heterozygote for the allele N (CSN1S1 F/N). Conversely, grouping the three genetic types, a high frequency of alleles associated with a low (CSN1S1 F ) and high (CSN1S1 A*,B* ) content of αs1 casein content in the milk, with CSN1S1 F ≥ CSN1S1 B* ≥ CSN1S1 A* , was observed (Table 1).
We compared the allele frequencies measured in the present work with those reported in different studies on goat populations reared in Italy and genotyped with the same or comparable techniques in this study. This comparison shows that these Lazio goat breeds have an intermediate genotype between the goat breeds reared in Northern Italy (Frisa, Orobica, Verzasca, Vallesana, Saanen, and Roccaverano), which are characterized by a high frequency of alleles associated with a null or low/intermediate αs1-Cn content. Instead, in the authocthonous breeds of Southern Italy, the alleles CSN1S1 B* and CSN1S1 A* are predominant (Table S2). The Napoletana goat breed is an exception within this panorama, and it stands out from the other breeds for the high frequency of the null allele CSN1S1 N and the CSN1S1 F allele (Table S2). This distinct genetic structure could be the consequence of geographic isolation after domestication. The Napoletana goat is mainly reared in the Lattari Mountains (the Campania Region), and it could have better preserved ancient alleles or variants rare or absent in other populations. On the contrary, in some breeds of Alpine origin (Saanen and Alpine Italian Chamois) the genetic selection of recent years is causing an increase in the frequency of strong and intermediate alleles at the expense of those of the weak and null [49].
For the remaining breeds, there are no genetic improvement actions. Consequently, the allelic frequencies remain somewhat stable over time, giving well-defined genetic structures to these breeds, primarily responsible for the particular chemical-physical, technological, and organoleptic characteristics of the milk produced and their derivatives.

Association Study
The polymorphisms of CSN1S1 affect not only the quantity of casein in goat milk but also its structural and nutritional characteristics (the diameter of the casein micelles, calcium content, fat, fatty acid profile, and urea level) [50][51][52][53][54][55], and technological properties (the coagulation parameters, cheese yields, and organoleptic properties) [22,56]. Moreover, a greater digestibility of goat milk containing αs1-Cn weak or null content has been shown [19].
In the present study, we have found an association between different genotypes at the CSN1S1 locus and some milk traits, namely the fat and protein yielded and fat, protein, solids-not-fat, and casein percentages, without having an effect on the milk yield. Similar results were observed when the goats were grouped into three clusters (strong, intermediate, and weak genotypes). This is in line with the literature being similar to those reported by [52], obtained by comparing goats with different CSN1S1 genotypes. These authors grouped the animals into two separate clusters called high (A, B, and C alleles) and low (F, G, 0, and E alleles). They found that milk produced by the low group has lower protein and fat contents than the high group, with no difference in the milk yield and lactose concentration. Moreover, Balia et al. [51] reported a lack of association between the CSN1S1 genotype and the milk yield in a flock of Sarda goats. Conversely, the milk protein and casein percentages were significantly affected by the genotype: the milk obtained by BB individuals was characterized by a high protein percentage than that of the AF and BF. High amounts of αs1-Cn expressed by the CSN1S1 BB genotype have also been reported in Cilentana goats without an effect on the milk protein and total casein concentrations [27].
Interestingly, goat breeds in this report show a trend to higher protein and casein percentages in homozygote B* goats than homozygote A* ones at the CSN1S1 gene. In support of that, Montalbano et al. [57] refer that quantifying by means of RP-HPLC B* genetic variant compared to A* show that the expression of this allele determines a higher content of αs1-casein in Girgentana goat milk.
A different degree of expression between the CSN1S1 A* and B* alleles may be a consequence of the presence of rare and not well-characterized A* alleles associated with a lower synthesis level, such as I (intermediate) or G (low) or 02 (null) [2]. The applied genotyping method by means of XmnI PCR-RFLP does not differentiate among these alleles. Another possible hypothesis is that there are individuals with B* allele variants, such as CSN1S1 B3 , with significant effects on the protein and casein percentages [58].
Furthermore, variations in the upstream region of the CSN1S1 gene may affect the protein expression and significantly affects the protein percentage [59][60][61][62]. In particular, a binding site of the Activator Protein (AP-1), known to be the critical third messenger for the target genes, regulated by extracellular mediators, and involved in the gene regulation of the mammary epithelial cells, as a response to prolactin, is affected by an A→G exchange at −175 bp in the bovine CSN1S1 promoter, associated to variations in the expression of the corresponding gene product [62]. Similarly, Ramunno et al. [7] refer that the mutation AJ504711:g653A>G seems to create an extra AP-1 binding motif in the proximal promoter sequence of the goat's CSN1S1 B* derived alleles. Therefore, it is possible to hypothesize that the mutation could be responsible for the observed expression level changes of goats' strong alleles (A* e B*). Currently, studies are ongoing to verify this hypothesis (Cosenza, unpublished data).

Conclusions
The opportunity to accurately characterize the genetic structure of loci of interest in livestock species allows us to better plan the breeding and selection activities. The new genotyping technique reported in this paper enables a more accurate characterization of the CSN2 locus in goats. Future aims include the development and application of an analysis protocol for each variant at the casein loci to rate their effect on milk traits.
Moreover, in this study, the autochthonous goat breeds of the Lazio Region have been genetically characterized for the first time at the quantitative alleles of calcium-sensitive caseins. These data are essential for the correct genetic management of these breeds to avoid the modification of their population's genetic structure and to guarantee the preservation of typical productions over time. Moreover, such breeds play two crucial functions, allowing the use of marginal areas, avoiding their abandonment, and preserving genetic variability and biodiversity.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/ani13020239/s1, Figure S1: (a) flock of the Bianca Monticellana goat breed that goes to pasture; (b) the Capestrina goat at recovery; (c) the Ciociara Grigia goat breed at the pasture; Figure S2: Goat CSN2 exon 7 nucleotide sequence (capital letters), plus the parzial intron flanking region (small letters). Primer sequences for AS-PCR reported by Ramunno et al. [36] are double underlined. Amino acid sequence is in uppercase and bold letters. Primer sequences for AS-PCR reported in this work are boxed. The transversion AJ011018.3:g.8913C>T responsible for the amino acid exchange, p.Ser166>Tyr, in the mature protein encoded by the CSN2 E allele is highlighted in gray. The CSN2 01 premature stop codon is symbolized by *; Table S1: Primer sequences and position used for AS-PCR; Table S2: Comparison of the allele frequencies for the CSN1S1 locus in different Italian goat breeds. References [63,64] are cited in the supplementary materials.

Institutional Review Board Statement:
This study was carried out following the recommendations of the European Council Directive (86/609/EEC) on the protection of animals. Ethical approval was not required in this study, as the milk and blood samplings were performed by a vet as part of a routine clinical visit.
Informed Consent Statement: Informed consent was obtained from the owners of the animals.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.