Plant material

In a previous study by ^{Rodríguez et al. (2016}), direct crosses were made, where the female parents were three sun poinsettias (Amanecer navideño, Belén and Juan Pablo) and the male parents were three shade poinsettias (Festival red, Ice Punch and Burgundy). From the progeny obtained, 39 F₁ plants were selected as the object of study for their different and outstanding morphological characteristics from their male parents and female parent, such as bract color, bract shape and appearance of the plant. Samples of 200 mg of young leaves were taken from each of the 39 plants, leaves were also taken from the six parents, were weighed, labeled and frozen at -20 °C until DNA extraction.

DNA extraction and quality

For DNA extraction, the method described by ^{Andrade et al. (2005}) was used, which is based on CTAB. DNA integrity was estimated by 1% agarose gel (ultrapure GIBCO BRL) electrophoresis, the gel was stained with 1 (g ml^-1 of ethidium bromide (Invitrogen^®) and subsequently observed in an ultraviolet light transilluminator with a photodocumenter (Syngene^® GVM20). The purity of the samples was determined by quantifying the amount of DNA with a NanoDrop 2000 spectrophotometer (Thermo scientific^®). The DNA was diluted to a concentration of 20 ng µl^-1 to be used in PCR reactions.

Fragment amplification

In a previous study ^{Rodríguez et al. (2016}), DNA amplification was performed with 68 random sequence primers of ten nucleotides from the OPA, OPB, OPC kit, plus 8 primers, of which, 10 were selected based on the quantity and brightness of the amplified RAPD fragments (OPA-07, 5’ GAA ACG GGT G 3’; OPA-11, 5’ CAA TCG CCG T 3’; OPA-12, 5’ TCG GCG ATA G 3’; OPB-07, 5’ GGT GAC GCA G 3’; OPB-08, 5’ GTC CAC ACG G 3’; OPB-09, 5’ TGG GGG ACT C 3’; OPB-11, 5’ GTA GAC CCG T 3’; OPB-20, 5’ GGA CCC TTA C 3’; OPC-06, 5’ GAA CGG ACT C 3’; OPC-13, 5’ AAG CCT CGT C 3’. The reaction components were as follows: 10 μl of dNTPs (5 μM of each dNTP, GIBCO BRL), 2.5 μl of PCR buffer (10x), 1.5 μl MgCl₂ (3 mM), 2 μl primer (10 pmol μl^-1), 0.3 μl (1.5 U) of native Taq DNA polymerase (Invitrogen), 4 μl DNA (20 ng μl^-1) adjusting to 25 μl with 4.7 μl sterilized deionized water.

DNA amplification was carried out using RAPD markers, which are based on the polymerase chain reaction (PCR) technique. The separation of amplified fragments was done by 1.5% (w/v) agarose gel (Invitrogen^®) electrophoresis, stained with ethidium bromide (10 μg l^-1). Electrophoresis was performed at 75 V per 4 h with TAE 1X. The documentation of the gels was carried out in the same way as in the quality of the DNA.

Data analysis

The size of the DNA fragments that were generated by the RAPD primers was obtained using the Gene tools Labworks 4^® program. The electrophoretic profiles generated by the parents and hybrids of each poinsettia variety were performed. The construction of the dendrogram was carried out to obtain the coefficient of similarity between progeny and parents, the data were analyzed based on the presence (indicated by 1) or the absence (indicated by 0) of amplified fragments with the selected decamer primers. Data were processed using the numerical taxonomy and multivariate analysis system (NTSYSpc 2.1). Clustering analyses were performed on the relationship of matrices with the linkage method of unweighted pair groups with arithmetic mean (UPGMA) (^{Avise, 1994}).

Identification of the progeny of Juan Pablo x Ice Punch and Juan Pablo x Festival red

With the primers used for DNA amplification of the crosses of Juan Pablo with Ice Punch and Festival red, monomorphic and polymorphic bands were obtained, 149 amplified fragments were obtained in total, of which 13% were polymorphic. The OPB-11 primer amplified the largest number of DNA bands with 23 amplified fragments and only 9% were polymorphic (Table 1), the primer that obtained the highest percentage of polymorphism was OPB-09, with 63%. The percentage of polymorphism ranged from 9 to 63%.

The size of the bands that identified the hybrid origin of the F₁ plants ranged from 446 bp to 2526 bp. The primers that allowed identifying the hybrid origin of the F₁ plants of the Juan Pablo x Ice Punch cross were those shown in Table 2, being the OPA-07 primer the one that most identified plants from the cross, and the plants that could be identified as hybrids with other primers were the plants 2 JP x Ip, 3 JP x Ip, 4 JP x Ip, 5 JP x Ip, 6 JP x Ip, 7 JP x Ip and 8 JP x Ip. This condition of similarity of bands with the male parent allowed corroborating that hybridization between the female Juan Pablo parent with the male Ice Punch parent was obtained.

The analysis of the conglomerate obtained (Figure 1) shows that, at a level of 0.26 of genetic distance, 6 groups form: 1) Juan Pablo and Festival red, 2) Ice Punch, 3) 2 JP x Ip, 4 JP x Ip, 3 JP x Ip, 6 JP x Ip, 7 JP x Ip, 8 JP x Ip,1 JP x Fr and 2 JP x Fr, 4) 3 JP x Fr, 5) 3 JP x Ip and 6) 1 JP x Ip. The group that concentrated more F₁ plants was group 3. The hybrids most similar to each other (with a coefficient of 0.15) are 1 JP x Fr and 2 JP x Fr, these are the ones with the lowest genetic distance. In contrast, the plant that does not share genetic similarity with any parent or progeny is the male Ice Punch parent (Figure 1).

The above could explain why in the Juan Pablo x Festival red cross, less progeny was achieved (3), since by sharing greater genetic similarity, there is less possibility of fertilization due to genetic closeness. In this regard, ^{Jiménez-Durán and Cruz-García (2011}) mention that the strong genetic control exercised by sexual incompatibility systems greatly restricts introgression with parents and crosses with other members of the progeny.

Figure 1 Relationship of the crosses generated on poinsettia, by Juan Pablo x Ice Punch (8 plants) and Juan Pablo x Festival red (3 plants) as determined by the UPGMA grouping analysis taking 149 RAPD fragments as a reference. 

On the contrary, the female Ice Punch parent is in a quadrant separate from its progeny and the other parents, showing a longer distance of separation between them because Ice Punch is the product of an interspecific cross between other poinsettias and there is greater genetic distance. The distribution obtained in the crosses generated by Juan Pablo x Ice Punch and Juan Pablo x Festival red through a two-dimensional graphic distribution gave rise to four more detailed groups: 1) the male Ice Punch parent with the crosses of Juan Pablo x Ice Punch 2, 3, 4, 6, 7 and 8; 2) the female Juan Pablo parent with the cross 5 of Juan Pablo x Ice Punch; 3) the male Festival red parent with the crosses 1 and 2 of Juan Pablo x Festival red and 4) the crosses 1 of Juan Pablo x Ice Punch and 3 of Juan Pablo x Festival red (Figure 2).

Figure 2 Graphic distribution of two male parents Ice Punch and Festival red, one female Juan Pablo parent, 8 plants generated in the Juan Pablo x Ice Punch cross and 3 plants generated in the Juan Pablo x Festival red cross. 

The RAPD markers failed to identify the progeny from the male Festival red parent, which may be due to the fact that hybridization between Juan Pablo and Festival red was not carried out, despite the fact that the hybridization process was taken care of so that they were not fertilized by other pollen from other plants or self-fertilized, there is a possibility that the plants are the product of self-fertilization and that is why they grouped together in group III, contrary to group I, where the male Ice Punch parent grouped with its progeny of Juan Pablo x Ice Punch, the uniformity of the behavior of this group and the identification with RAPD markers shows that hybridization was carried out, evidencing the genetic identity of the six progeny studied.

Identification of the progeny of Amanecer navideño x Ice Punch and Amanecer navideño x Festival red

The amplification of the RAPD primers used in the study of the crosses obtained from Amanecer navideño x Ice Punch and Amanecer navideño x Festival red revealed the presence of 146 fragments with a percentage of polymorphism of 34% (Table 3). The OPB-07 and OPB-20 primers were the ones that generated the greatest amplification of fragments (20) with polymorphism of 35% and 50% respectively, although the OPB-11 primer generated less fragments, it had 56% polymorphism, so the percentage of polymorphism varied from 27 to 56%.

The fragments that identified the progeny of the crosses of Amanecer navideño x Ice Punch and Amanecer navideño x Festival red ranged from 661 to 2455 bp. The primer that generated the largest number of fragments that identified hybrid plants was OPA-07, the percentage of polymorphic information of this primer was 33%. The primers that did not detect any hybrid obtained from the crosses were OPC-06 and OPC-13 (Table 3).

The plants that could be identified as hybrids with most of the primers were 2 An x Ip, 3 An x Ip, 4 An x Ip, 5 An x Ip, 6 An x Ip, 7 An x Ip, 2 An x Fr and 3 An x Fr. The above suggests that the crosses made of Amanecer navideño x Ice Punch were effectively achieved, since it was possible to identify its progeny with respect to the male Ice Punch parent, except for the plant 1 of this cross that did not show any fragment in common with the male Ice Punch parent.

The dendrogram generated (Figure 3) from the data obtained with the ten RAPD primers showed the genetic closeness between plant 5 An x Ip and 6 An x Ip, as well as plant 3 An x Ip and 4 An x Ip. Six groups formed at a genetic distance coefficient of 0.17: 1) Amanecer navideño with Ice Punch; 2) 3 An x Ip, 4 An x Ip, 7 An x Ip; 3) 2 An x Ip; 4) male Festival red parent, 1 An x Ip; 5) 2 An x Fr and 3 An x Fr; 6) 5 An x Ip, 6 An x Ip and 1 An x Fr.

Figure 3 Relationship of the crosses generated on poinsettia, by Amanecer navideño x Ice Punch (7 plants) and Amanecer navideño x Festival red (3 plants) as determined by the UPGMA grouping analysis taking 146 RAPD fragments as a reference. 

The plants most similar to each other (with a coefficient of 0.12) are 5 An x Ip and 6 An x Ip; as well as the progeny 3 An x Ip and 4 An x Ip, these are the ones that present the least genetic distance, in the same way, Amanecer navideño and Ice Punch parents share genetic similarity to a distance coefficient of 0.15. In contrast, the plant that does not share genetic similarity with its progeny or another parent is the male Festival red plant (Figure 3).

The conglomerate shows that the ten F₁ plants of the crosses of Amanecer navideño x Ice Punch and Amanecer navideño x Festival red joined different groups, which shows genetic closeness between the three according to what was formed in the three groups formed: 1) the male Ice Punch parent; 2) the crosses of Amanecer navideño x Ice Punch 2, 3, 4 and 7; and 3 ) the female Amanecer navideño parent, the male Festival red parent with crosses 1, 5 and 6 Amanecer navideño x Ice Punch and the crosses 1, 2 and 3 of Amanecer navideño x Festival red (Figure 4).

Figure 4 Graphic distribution of two male parents Ice Punch and Festival red, one female Amanecer navideño parent, 7 plants generated in the Amanecer navideño x Ice Punch cross and 3 plants generated in the Amanecer navideño x Festival red cross. 

Identification of the progeny of Belén x Ice Punch, Belén x Burgundy and Belén x Festival red

The amplification of the RAPD primers used in the study of the crosses obtained from Belén x Ice Punch, Belén x Burgundy and Belén x Festival red, revealed the presence of 113 fragments with a percentage of polymorphism of 26% (Table 2). The OPC-13 primer was the one that generated the highest amplification of fragments (14) with a polymorphism of 36%, but the primer that gave the greatest polymorphism was OPC-06 with 38% and 13 amplified fragments, the percentage of polymorphism varied from 10 to 38%.

The primers generated monomorphic marker fragments which were invariable in all the study plants, and these could not identify the parents with the hybrid progeny, the fragments did not show differences even between the same parents. The dendrogram generated (Figure 5) from the data obtained with the 10 RAPD primers showed the formation of 11 groups at a genetic distance coefficient of 0.18, the groups formed as follows: 1) the female parent Belén with Ice Punch; 2) the male Burgundy parent; 3) the male Festival red parent; 4) 2 Be x Bur; 5) 9 Be x Fr, 10 Be x Fr and 13 Be x Fr; 6) 11 Be x Fr; 7) 12 Be x Fr; 8) 1 Be x Ip, 1 Be x Bur, 4 Be x Fr, 5 Be x Fr, 2 Be x Fr, 3 Be x Fr; 9) 1 Be x Fr, 6 Be x Fr; 10) 7 Be x Fr, 8 Be x Fr; and 11) 3 Be x Bur.

Figure 5 Grouping of male parents (Ice Punch, Festival red and Burgundy) and the relationship of the crosses generated with Belén, as determined by the UPGMA grouping analysis taking 113 RAPD fragments as a reference. 

The hybrid plants most similar to each other (with a coefficient of 0.09) were 1 Be x Bur and 4 Be x Bur, as well as 2 Be x Fr and 3 Be x Fr, Ice Punch and Belén parents share genetic similarity at a distance coefficient of 0.16 (Figure 5). The graphic distribution shows that the 18 F₁ plants of the crosses of Belén x Ice Punch, Belén x Festival red and Belén x Burgundy are genetically different with respect to the four groups formed: 1) the male Ice Punch parent; 2) the hybrid 11 of Belén x Festival red; 3) the hybrids 1, 2, 3, 4, 5, 6, 7, 8 and 9 of Belén x Festival red and 1, 3, 4 of Belén x Burgundy; and 4) the crosses of the female Belén parent with the male Festival red parent and its crosses 1, 2, 3, 4, 5, 6, 7, 8 and 9; Belén x Ice Punch, and cross 1, the hybrids of Belén x Burgundy 1, 4 and 5 (Figure 6).

Figure 6 Graphic distribution of the male parent Belén x Ice Punch 1 plant generated, Belén x Festival red 3 plants generated, and Belén x Burgundy 14 plants generated in the crosses. 

The primers selected for the parents Juan Pablo, Amanecer navideño and Belén and the plants originated from the poinsettia crosses of Juan Pablo x Ice Punch, Juan Pablo x Festival red, Amanecer navideño x Ice Punch, Amanecer navideño x Festival red, Belén x Ice Punch, Belén x Festival red and Belén x Burgundy, produced 408 RAPD fragments in total, of which 109 fragments were polymorphic with 27% polymorphism with 10 primers, these allowed identifying the F₁ plants of Juan Pablo x Ice Punch, Juan Pablo x Festival red, Amanecer navideño x Ice Punch, Amanecer navideño x Festival red with 26 marker fragments, the crosses with Belén could not be identified with the 10 primers, since they generated marker fragments; however, these marker fragments did not allow identifying any hybrid, 11 marker fragments were produced, which could be evaluated with other primers.

This shows the discriminating capacity of RAPDs for the identification of poinsettia cultivars. The results of this research coincide with what was done by ^{Jing-Tian et al. (1997}), who identified poinsettia cultivars using RAPD markers, performed the amplification of 60 primers and obtained 69% polymorphism and conclude that their results indicate that RAPDs are efficient for the identification of Poinsettia cultivars and for the determination of genetic relationships between cultivars.

In another AFLP evaluation study in Poinsettia: polymorphism selection, analysis and identification of cultivars, they evaluated 81 poinsettia cultivars using 41 polymorphic AFPLs, of which they identified only 60 poinsettia cultivars, the other 21 were not identified (^{Parks and Moyer, 2004}). In a study of comparison of RAPD and AFLP as methods of genetic identification of vines based on the study of fragments, they studied 57 varieties of vine with 80 primers, in their RAPD study, they mention that this result is remarkable since a correspondence between the genealogy and the information provided by the molecular analysis is verified and they conclude that both methods allowed differentiating all the cultivars analyzed (^{Narváez et al., 2000}).