Genetic diversity and geographic population structure of bovine Neospora caninum determined by microsatellite genotyping analysis

Abstract

The cyst-forming protozoan parasite Neosporacaninum is one of the main causes of bovine abortion worldwide and is of great economic importance in the cattle industry. Recent studies have revealed extensive genetic variation among N. caninum isolates based on microsatellite sequences (MSs). MSs may be suitable molecular markers for inferring the diversity of parasite populations, molecular epidemiology and the basis for phenotypic variations in N. caninum, which have been poorly defined. In this study, we evaluated nine MS markers using a panel of 11 N. caninum-derived reference isolates from around the world and 96 N. caninum bovine clinical samples and one ovine clinical sample collected from four countries on two continents, including Spain, Argentina, Germany and Scotland, over a 10-year period. These markers were used as molecular tools to investigate the genetic diversity, geographic distribution and population structure of N. caninum. Multilocus microsatellite genotyping based on 7 loci demonstrated high levels of genetic diversity in the samples from all of the different countries, with 96 microsatellite multilocus genotypes (MLGs) identified from 108 N. caninum samples. Geographic sub-structuring was present in the country populations according to pairwise F(ST). Principal component analysis (PCA) and Neighbor Joining tree topologies also suggested MLG segregation partially associated with geographical origin. An analysis of the MLG relationships, using eBURST, confirmed that the close genetic relationship observed between the Spanish and Argentinean populations may be the result of parasite migration (i.e., the introduction of novel MLGs from Spain to South America) due to cattle movement. The eBURST relationships also revealed genetically different clusters associated with the abortion. The presence of linkage disequilibrium, the co-existence of specific MLGs to individual farms and eBURST MLG relationships suggest a predominant clonal propagation for Spanish N. caninum MLGs in cattle.

Figure 1. Clustering of N . caninum country populations based on Principal Component Analysis (PCA) and Neighbor Joining (NJ).

(A) Graphics represent the genetic relationships between the MLGs from each country population based on a covariance matrix. Colours indicate geographic origin (see legend). The proportion of variation in the dataset for each axis is indicated in parentheses. (B) Unrooted NJ tree inferred from allele shared genetic distances. Each tip represents a single MLG. Woldwide MLGs are identified by isolate name. Colour of circles in terminal branches indicates geographic origin (see legend). Percentage bootstrap values were generated from 1,000 replicates. Bootstrap values ≥70% are shown in black circles. Scale bar represents branch lengths. NJ tree based in Cavalli-Sforza distances depicted similar topologies (data not shown).

Figure 2. Relationships among the N . caninum MLGs (n = 82) estimated by eBURST analysis.

Each complete MLG is represented by a dot and the genotype number assigned in Table S1 (Numbers 1-11 represent worldwide MLGs; 12-64 Spanish MLGs; 65-80 Argentinean MLGs; 81-88 Scottish MLGs; and 89-95 German MLGs). MLG dots were also coloured according to their geographical origin (see legend). The dot diameter is proportional to the number of samples with identical MLG (see legend). Single locus variants (SLV) are connected by black lines and double loci variants (DLV) by blue lines. MLGs clusters (n=9) are represented. Main MLGs clusters (n=3) are showed by clear circles. Number and name in the squares identifies the herd origin and worldwide isolates, respectively (see Table S1). Note that singletons were excluded from the snapshot representation (worldwide n=7; Spanish n=7; Argentinean n=2; Scottish n=3; and German n=2).