Desenvolvimento de mapas genéticos e físicos com base em SNPs em Psidium guajava e triagem in silico de loci microssatélites de goiabeira para transferibilidade em Eucalyptus globulus

Abstract

Guava tree (Psidium guajava) is worldwide known for its easy cultivation, nutritional richness and consumption versatility. Although this crop is expanding commercially, it is still little studied from a molecular point of view. Thus, it is important to promote studies that enable marker assisted selection in guava. Given that, this study aimed to 1) develop and compare P. guajava genetic and physical maps based on SNPs from Eucalyptus and 2) propose BLAST for in silico screening of microsatellite sequences (SSR) from P. guajava to E. globulus in order to guide transferability studies. Guava’s plant material was originated from the cross between Pedro Sato × Goiabeira Roxa (PSR), available at Embrapa Semiárido, Petrolina – PE, Brazil. Young and healthy leaves, from both species, were collected, properly identified, and put in styrofoam for transportation to the laboratory where samples were extracted with a modified CTAB 2x protocol, including a sorbitol pre-wash. DNA quantification was performed by spectrophotometry, using a Microdrop plate in the absorbance range of 260nm. After quantification, samples were diluted to work concentration of 60 ng/μL (P. guajava) e 20 ng/μL (E. deglupta) and stored at -20ºC. Genotyping of the 112 PSR was performed with Euchip60K from Eucalyptus. Linkage analysis was performed using JoinMap 4.0, LOD score from 5 to 12 meanwhile GACD analysis a LOD score of 8. A physical map was ordered in guava chromosomes considering Eucalyptus SNPs (query) alignment to guava’s genome (subject), estimating marker position via Blastn with a cut off e-value < E-10. JoinMap 4.0 generated a map of 1.405,2 cM and GACD a map of 1.392,7 cM. Both maps presented linkage groups with segments from several chromosomes when compared to the physical map, indicating limitations. GACD showed a greater limitation in comparison to JoinMap 4.0 because it divides markers according to their parental origin. The physical map generated with Blastn, e-values ranging from 8xE-10 to 1.15xE-26 covered 434. 88Mb, with mean distances of 0.6Mb, being a reference for mapping and QTL estimations in guava. For the in silico screening 23 P. guajava SSR clones (query) were aligned against E. globulus genome (subject), using Blastn and e-value < 1E−20. Also 140 primers made available by GuavaMap were analyzed based on e-value < 1.7 and the distance between forward and reverse sequences (a maximum of 300 nucleotides distance). 39% of the 23 SSR clones showed significant alignments with sequences mean identity of 87%. Regarding GuavaMap loci only three, 2.1%, presented significant alignments. SSR loci with significant BLAST results also amplified in polyacrylamide gel meanwhile the loci that did not show “hits”or e-value on Blastn did not produce amplicons in vitro. Therefore, in silico screening showed to be effective in reducing costs, time and in orienting transferability studies between these species.