Coat genetic markers of the domestic cat Felis catus ( Felidae ) from southwestern Colombia Marcadores genéticos del pelaje del gato doméstico Felis catus ( Felidae ) del suroccidente colombiano

Objectives. Establish the genetic profiles of cats from 12 neighboring municipalities in southwestern Colombia, in a town course from Pereira-Popayán. Estimate the degree of diversity, genetic structure, and quantify gene flow. Materials and methods. Were inventoried the phenotypic markers present in the pigmentation and structure of the coat of 1482 cats of the municipalities surveyed. Based on these phenotypic frequencies, allele frequencies, heterozygosity, Hardy-Weinberg equilibrium, F statistics and Nei genetic distances were calculated. A comparison was also made between genetic and geographic distance matrices to determine if there was a significant association between the two. Results. With the genetic profiles of the populations we estimated the degree of diversity. We found the populations in equilibrium for the S autosomal locus and for the O sex-linked locus. We found a low level genetic structure, and it was determined that there was no significant correlation between the genetic and geographic distance matrices among populations. Conclusions: These findings can be explained on the basis of the processes of human displacement for this region, due to the fact that the establishment of feline populations in these municipalities originated during the same historical period. Identical genetic profiles are shared as a result of colonization events, and due to possible continued migration among these populations.


INTRODUCTION
During the past century genetic coat markers in domestic cats Felis catus (Felidae) have become a useful tool to study the genetics of populations, since they have different types of heredity and genic actions such as epistasis and incomplete dominance that enriches the research.Recently, the genetic profiles of numerous American (1-4), European (5)(6)(7)(8) and Asian (9) populations have been published.Based on an inventory of the allelic frequencies of the various loci that affect the color, pattern and length of the coat, we have been able to establish a hypothesis of historical migration through the world.
Genetic profiles of cat populations have served to assess the hypothesis of historical migration of colonial populations of cats during transcontinental events.We have studied the possible change in allele frequencies in cat populations among European countries with regard to their transoceanic colonies (2,10,11).In each case a differentiation has been observed between the allelic frequencies in populations of (European) origin and the resulting (colonies).
This study considers a hypothesis of historical migration in a more reduced geographical space in relation to the transoceanic migrations of European countries.Populations were studied from Pereira to Popayán, going through the Valle del Cauca, establishing the allelic frequencies in 12 municipalities and contrasting it with historical human migrations in Colombia.This region, where the study is conducted, is of great national economic development, due to its intense agricultural productivity, the industrial settlement and its large trade flow because of its proximity to the main Pacific port, Buenaventura.
All these factors have turned the biogeographical region of the Cauca River into a melting pot of various Colombian cultures that have probably affected the genetic profiles of cats in migratory events.Finally, the objectives of this investigation were: to identify, through coat genes, the Peñuela -Geneticmarkers of cat coats in southwestern Colombia esta investigación fueron: caracterizar, mediante genes del pelaje, poblaciones de gatos domésticos de 12 municipios del suroccidente colombiano, estimar el grado de estructura de las poblaciones, y determinar si existe una correlación entre las distancias geográficas y genéticas para contrastar los resultados frente a los procesos históricos de migración humana.

MATERIALS AND METHODS
Sampling.A total of 1482 individuals were photographically sampled in 12 southwestern Colombian municipalities, encompassing three states covered by the region studied of the biogeographic valley of the Cauca river (Figure 1).Their genetic profiles were inventoried again, to determine if the cat populations of Cali and Popayán have changed their allelic frequencies throughout the years.Therefore, the number that goes with these cities indicates whether they have been in the first, second or third inventory.
In small municipalities, the sampling consisted of walks around town to enquire at each home, whether they had feline pets.For large cities the sampling was done by dividing the city into grids and the sampling sites were chosen at random.In households where the answer was affirmative, a request was made to photograph the animals and proceed to the collection of metadata such as sex, age (in months) and whether they were sterilized.In some municipalities visits were made to shelters or foundations sheltering abandoned pets.The corresponding photographic sampling was conducted.However, for these Peñuela -Geneticmarkers of cat coats in southwestern Colombia cromosómico de estos marcadores genéticos con base en la información publicada (Figura 2).cases no metadata could be collected due to the lack of said data.

Análisis de datos. El cálculo del equilibrio
Identifying phenotypic markers.The genetic nomenclature used follows the Committee on Standardized Genectic Nomenclature for Cats (12) proposal, except for the chromosomic positions that make up the Tabby patterns, for which we followed Eizirik et al (13).The genes used for the genetic inventory of populations are listed in (Table 1).
Research in the last decade has been able to identify the chromosomal positions of a great portion of the genes studied (14)(15)(16)(17)(18)(19)(20)(21)(22), reason why the classical phenotypic work has been transformed into a detailed inventory of genes, where chromosomes are converted into sampling units; in this article we develop a chromosomic ideogram of these genetic markers based on published information (Figure 2).The expected heterozygosity was calculated as (1-Σp i 2 ) or (2pq) and the standard deviation of the allelic frequencies as (√(pq⁄2n)).We calculated the Wright index of F ST differentiation from the O, a,Ta(b), d, S, I, W, l and c s genes (Table 2).The matrices obtained were computed in the Mega6 program (23) to generate dendrograms from Neighbor Joining algorithm.Dendrograms were also used to compare the study populations with Colombian populations previously inventoried, such as: Bogotá, Bucaramanga, Capurganá, Duitama, Ibagué, Pasto, Leticia, Leticia-Tabatinga and the previous samples from Cali and Popayán (Table 3).This table has the allelic frequencies and standard deviations of 11 morphological genes that affect the coat traits of F. catus, previously published in Colombia and the source data is listed in parenthesis.To test the hypothesis that the genetic distances of populations are correlated with geographic distances, the Mantel test was performed with 999 repetitions in the Gen ALEx program (24), using the F ST matrices and Nei genetic distances (D) versus the geographical distance matrix calculated from the latitude and longitude of each municipality.

RESULTS
Table 2 shows the average allele frequencies of the loci studied with their standard deviations.The highest O allelic frequencies oscillated between 0.2979 and 0.1573; the allele a between 0.8358 and 0.6922.The Ti(A) allele was only detected in a heterozygous individual from Popayán-2 (0.0157).The Ta(b) allele was not found in any individual of Bugalagrande.The d allele showed a Los valores obtenidos en las matrices F ST y las distancias genéticas (D) de Nei no presentaron correlación significativa con la matriz de distancia geográfica obtenida a partir de datos de latitud y longitud, prueba de Mantel con 999 repeticiones, F ST (r 2 = 0.03, p = 0.08) y (D) de Nei (r 2 =0.00, p=0.380).

El dendrograma generado por el algoritmo
Neighbor joining con base en los valores de F ST de las poblaciones inventariadas (Figura 3a), muestra una agrupación entre las ciudades range of frequencies between 0.5145 and 0.3753, and the S allele between 0.4225 and 0.2549.
The I allele showed lowest frequencies between 0.0465 and 0.0109.In addition, the W allele had low frequencies between 0.0282 and 0.0075.The l allele showed frequencies between 0.5863 and 0.3462; the c s allele between 0.45969 and 0.2596.The M allele was only observed in a heterozygous individual from Cartago (0.0026) and other from Jamundí (0.0052).Finally, the b allele was not found, the B allele was fixed in all populations.
In each locality and in a mixed group, the Hardy-Weinberg equilibrium was estimated for the loci S and O.All assed populations did not show deviations from the Hardy-Weinberg equilibrium (Tables 4 and  5).However, to the O locus, the population mixed showed significant deviations from equilibrium (Table 5).This sample shows a homozygous excess and heterozygous deficiency, phenomenon known as the Whalund effect.This effect refers to the heterozygosity reduction in the total population caused by the structure of the populations.To know, whether two or more subpopulations have different allelic frequencies then the total heterozygosity it´s reduced, although subpopulations remain in the Hardy-Weinberg equilibrium.
The values obtained in the F ST matrices and the Nei genetic distances (D) showed no significant correlation with the geographical distance matrix obtained from latitude and longitude data, Mantel test with 999 repetitions, F ST (r 2 = 0.03 , p=0.08) and (D) of Nei (r 2 =0.00 , p=0.380).Por fuera de esta agrupación pero cercana a ella se encuentra Tuluá, la cual se desprende de Cartago, ésta última de Piendamó y Bugalagrande.Cali-3 presenta el mayor grado de diferenciación ubicándose en una rama distante por fuera de la agrupación y Jamundí la acompaña en su separación pero no forma agrupación sólida con Cali-3.

DISCUSIÓN
Se debe tener cuidado con estos registros dado que se obtuvieron bajo condiciones domésticas y probablemente difieran de las condiciones naturales The dendrogram generated by the Neighbor joining algorithm, based on the F ST values of the inventoried populations (Figure 3a), shows a grouping among the cities of Buga, Palmira, Obando, Pereira, Santander de Quilichao and Popayán-2.Out of this grouping, but close to it, we find Tuluá, which in turn is part of Cartago, the latter of Piendamó and Bugalagrande.Cali-3 exhibits the highest degree of differentiation placing itself on a distant branch outside the grouping, and Jamundí accompanies it in its separation, but does not form a solid grouping with Cali-3.
The dendrogram generated with the same algorithm, but which compares the populations under study with other Colombian populations (Figure 3b), shows a solid grouping among the populations studied, including the previous genetic profiles of Cali-1, Cali-2 and Popayán-1.
The other Colombian populations that do not belong to the biogeographical valley of the Cauca River shifted even further away from this grouping.And Leticia and Leticia-Tabatinga finally appear along with Capurgana with very different genetic profiles.

DISCUSSION
Care must be taken with these records since they were obtained under domestic conditions and are likely to differ from natural conditions or those lacking domestic care.However, the limited control by human beings over the choice of partners and the generation of litters with simultaneous fatherhoods, dramatically increase the genetic flow within populations, inbreeding reduces and increases panmixia.
The allelic frequencies of the twelve loci studied were very similar among the 12 populations (Table 2).The inventory of the highest allelic frequencies in dominant genes of low-penetrance such as: Ti(A), W, M in historically older populations, could genetically turn them into candidates as source populations of others.
The elevated variation of the allelic frequencies present in the Ta(b) locus singles it as a useful locus in the differentiation between populations, followed by the O, d and l loci.The Ta(b) locus has been previously catalogued as one of the most heterogeneous markers, since its frequencies have helped differentiate populations of British origin from those of Iberian origin (10).
In this study we report the absence of the b allele in all populations evaluated, reason why the B allele is fixed in.Frequencies for b are reported in (4), in the subsampling done in southern Cali.However, this datum was an incorrect identification of an individual (A.Peña Com.Pers.).Therefore, no study has reported the presence of this allele in Colombian populations.
Like Peña et al (4), a decrease was found in the allelic frequencies of a and d for the population of Cali with respect to the first sampling performed by Ruíz-García y Álvarez (2).Both Tb(a) and I increased their allelic frequencies, agreeing with what was reported by Peña et al (4).The relevant differences between this study and Peña et al (4) are found in the values of O (0.1574 versus 0.2481), l (0.3462 versus 0.5165) and c s (0.2826 versus 0.4383).The allelic frequencies reported here for Cali-3 are more similar to those reported for Cali-1 according to Ruíz-García and Álvarez (2).Peña et al (4) justifies that the increase in these frequencies in Cali-2 can be attributed to the human selection of more attractive features (orange cats, with long hair or Siamese cats).We agree with this hypothesis, since the differences between the frequencies of these loci in Cali-2 and Cali-3 are surely due to the sampling design, since for Cali-2 we only sampled individuals in their homes, while for Cali-3 we sampled individuals in homes and shelters for abandoned animals.In contrast to Cali-2 Peña et al ( 4), the Peñuela -Coat genetic markers of cats in southwestern Colombia siendo la frecuencia alélica muy baja (0.0039), al igual que el individuo heterocigoto reportado en este estudio en Popayan-2 (0.0157), resultados que concuerdan con nuestra hipótesis de que el inventario de las frecuencias alélicas en genes dominantes de poca penetrancia como: Ti(A), W, M en poblaciones históricamente más antiguas, pueden candidatizarlas genéticamente como poblaciones fundadoras de las otras.
Según las pruebas de Mantel realizadas a la matriz F ST de Wright e Identidad genética (D) de Nei, ninguna de éstas se correlaciona con la matriz de distancia geográfica calculada a partir de latitud y longitud.Esto indica que las poblaciones actuales no se asemejan genéticamente entre sí por proximidad geográfica, lo que deslegitimiza la hipótesis de que allele frequency for Ti(A) in Cali-3 is not reported here.It is not surprising, since only an individual in heterozygous state was inventoried, with a very low allele frequency (0.0039), and also the individual heterozygous reported in this study in Popayán-2 (0.0157).These results are consistent with our hypothesis that the inventory of the allelic frequencies in low-penetrance dominant genes such as: Ti(A), W, M in historically older populations, could genetically turn them into candidates as source populations of others.
It has been widely discussed that the presence of the Ti(A) allele in the continent, is due to the fact that there were trade routes between the ports of America in the Pacific and Asia, since the Ti(A) allele has the highest frequencies in southeast Asia (Southern coast of China, 0.20; Calcutta 0.37, Bombay 0.15, Colombo 0.30) (10,11).This would explain the presence of the allele in the population of Popayán, city that was one of the administrative centers of the Spanish Reign, and is also near the Pacific Ocean.
In the populations studied, we validated the existence of the Hardy-Weinberg equilibrium for the S locus and for O the locus using the method described by Christensen (25).(Tables 4 and  5), reason why we confirmed the stability of the allelic frequencies of these loci in the populations studied.Populations with demographic conditions with elevated effective numbers have proven to be highly stable over time (11).
The metapopulation(all mixed populations) exhibited no equilibrium in the O locus.Even though an autosomal locus such as S exhibits population balance, the calculation of the balance of the genes linked to sex is more sensitive due to the greater number of categories used by Chi-square.For the O locus, the mixed population did not behave as a panmictic unit, evidencing Whalund effect, heterozygous deficit or homozygous excess, indicating any degree of structure or population subdivision.The subjacent causes of this subdivision in the population could be geographic barriers at the gene flow followed by genetic drift in the subpopulations.
In comparison to other populations of this study, Cali-3 obtained the highest values of F ST differentiation.This difference could be caused by permanent human migration, since this city is the current economic and political center of the region.According to the Mantel tests conducted on Wright's F ST matrix and on Nei's genetic identity (D), none of these correlates with the geographical distance matrix calculated from latitude and longitude.This shows that current populations are not genetically similar among flujo genético constante por la cercanía geográfica haya sido el causante de las similitudes en las frecuencias genéticas del total de poblaciones.
El dendrograma Neighbor joining para los valores de F ST , detecta la formación de grupos muy sólidos entre las ciudades.En el primero, se observa una proximidad entre la mayoría de las ciudades a excepción de Cali-3, ésta por ser la ciudad que ha recibido históricamente la mayor migración de personas, pudo haber alterado de forma considerable su perfil poblacional frente a las otras poblaciones (Figura 3a The historical account that gave rise to the cities of the Colombian southwest enables us to understand the great genetic similarity among the cat populations.Here the hypothesis of a common genetic population as the origin of the other populations, finds strong foundations thanks to the history of cities of the region.The municipalities of Popayán, Cali and Cartago were the territorial administrative centers of Spain during the colonial era in the biogeographical valley.By being the only population present at the time, the flow of inhabitants and their pets must have been permanent, added to the fact that these municipalities belonged to the same government. The territorial flourishing of these populations and their constant communication through the Cauca river channels, could have allowed the establishment of a similar genetic profile among the source populations that would give rise to the rest of subsequent populations.This hypothesis could find support in the inventory of the highest dominant allele frequencies in low penetrance genes such as: Ti(A), W, M, in historically older populations.Thus the allele Ti(A) only was inventoried in heterozygous state in the municipality of Popayán-2, where the highest W frequencies of were also found.The M allele was only found in the municipalities of Jamundí and Cartago, obviously in heterozygous state, due to the lethal condition of its homozygosity.
The inventory of these allelic frequencies offers a greater genetic diversity to these populations, because it sets them up to be strong candidates as mother populations.
Let us remember that (2,10,11) group, in their analysis, the populations of Cali, Popayán, Buenos Aires, California and Colorado in the same gene pool.This grouping reveals the connection through maritime trade routes in the Pacific, reason why they are genetically similar and have a common genetic origin of Hispanic origin (10).
The genetic profiles of populations after the colonial era were probably caused by geographically nearby cities, which is supported in some groupings formed by dendrograms.Cali's genetic distancing can be explained by the events of the past centuries.This city became the engine of economic development in the region, because of its proximity to the port of Buenaventura.This economic prosperity really contributed to the massive migration of people from all over the region, which will certainly affected the genetic profile of the cat populations, but did not markedly differentiate it from the original genetic block.
It is clear that the origin of the genetic profiles of southwestern Colombian domestic cats presents a common historical origin that has generated little differentiation over time, and that genetically separates them from the remaining Colombian populations.The rapid population growth of the species and the minimal human control over their reproduction greatly increases gene flow and increases panmixia, key conditions to keep to a minimum the variation in genetic frequencies.

Figure 1 .
Figure 1.Map of southwestern Colombia depicting the locations of the study and their sample size.

Figure 2 .
Figure 2. Karyotype of the domestic cat Felis catus, the chromosomal positions of the alleles were arranged based on the previously published information.

Figure 3 .
Figure 3. Dendrograms generated by the Neighbor joining algorithm on the basis of the values of F ST for nine genes of F. catus, (a) in 12 municipalities in southwestern Colombia and (b) among populations of this study and Colombian populations previously inventoried.

Table 2 .
Allelic frequencies and standard deviations of 12 morphological genes that affect coat characteristics of the F. catus in 12 municipalities in southwestern Colombia.

Table 3 .
Allelic frequencies and standard deviations of 11 morphological genes that affect coat characteristics of the F. catus, previously published for Colombia.

Table 4 .
Chi-square to test the Hardy-Weinberg equilibrium hypothesis (Eq.H-W) in the S locus of F. catus in 12 municipalities in southwestern Colombia, n corresponds to the number of individuals surveyed for locus in each population.* X 2 tabulated (2; 0.05) = 5.991.

Table 5 .
Chi-square to test the Hardy-Weinberg equilibrium hypothesis in the O locus of F. catus in 12 municipalities in southwestern Colombia, n corresponds to the number of individuals sampled.* X2 tabulated (4;0.05)=9.488.
(26)ng the colonial era, Santiago de Cali and Cartago were part of the government of Popayán, which in turn was part of the Royal Audience of Quito(26).Guadalajara de Buga was founded in 1554, but the city's location changed 4 different times, until it finally settled in its current location.Tuluá was founded in 1539, and Bugalagrande in 1662, on land where the city of Buga had previously stood.Palmira was founded between 1680 and 1705 by the tenth century signed in Guadalajara de Buga.The transfer of Cartago to its current location on the plains of Valle del Cauca took place in the year 1691.In 1863, the ruins of the old Cartago were reconquered and it gave birth to the city of Pereira.The repopulation of these lands was greatly influence by the Antioquia colonization, as well as the foundation of Obando in 1890.