In Mexico, poinsettia (Euphorbia pulcherrima Willd. ex Klotzsch) is cultivated in backyards and gardens, as an improved species, and is located as a wild plant from Northwestern Mexico to the South of Guatemala (^{Trejo et al., 2012}; ^{Canul et al., 2014}). Poinsettia plants cultivated in gardens and backyards are known as "nochebuena de sol"; they are used as medicinal and cut flowers plants, and for decorating public spaces (^{Colinas, 2009}; ^{Canul et al., 2014}). There are no official data about the cultivated area with "nochebuena de sol"; in 2010, in the state of Morelos, they were cultivated around of 20,000 m² for regional, state and national markets, and without trade agreements (^{Galindo et al., 2012}).

In contrast, poinsettia, as an improved species (hereinafter referred as commercial poinsettias), is one of the most important ornamental plants worldwide. According to the United States of America flower production records, poinsettia sales in 2013 were $146 million dollars (^{USDA, 2014}). In the same year, the production value in Mexico was over $416 million pesos (^{SAGARPA, 2014}).

Wild poinsettia plants have been under a constant selection pressure, so they have developed some traits that add value as a genetic resource (^{SINAREFI, 2010}) with potential for research and marketing. Therefore, it is necessary to protect and incorporate this species to genetic improvement programs (^{Canul et al., 2010}). In this regard, INIFAP at the Zacatepec Experimental Field is developing a genetic improvement program for "nochebuena de sol" and wild poinsettia plants, aimed to obtain Mexican varieties with commercial and phytosanitary features that benefit farmers.

Commercial poinsettias are regular hosts for both Poinsettia mosaic virus (PnMV) and Poinsettia latent virus (PnLV) (^{aus dem Siepen et al., 2005}). Even though no symptoms have been reported on plants infected by PnLV (^{aus dem Siepen et al., 2005}) so far, it is known that PnMV induces bracts distortion, mosaic and leaf mottling, which cause important marketing losses (^{Lebas, 2007}).

In Mexico, research about phytopathogenic virus in "nochebuena de sol" and wild poinsettia plants is recent. There are two reports where the presence of such pathogens was associated with symptoms of chlorosis, mosaic, variegation and leaf deformation. In 2012, results of a preliminary test using immune strips showed the presence of Cucumber mosaic virus (CMV), Tobacco mosaic virus (TMV) and Tomato spotted wilt virus (TSWV). The presence of CMV and TMV was later confirmed using DAS-ELISA (^{Ocampo, 2012}). In 2013, TMV was detected using DAS-ELISA in "nochebuena de sol" plants collected in Morelos, Guerrero, Mexico City, Michoacán, State of Mexico, Puebla, Veracruz, Oaxaca and Nayarit. The presence of this virus was confirmed by RT-PCR in plants collected in the State of Mexico and Nayarit. PnMV was also detected in "nochebuena de sol" plants collected in Michoacán, State of Mexico, Veracruz and Oaxaca; these results were further confirmed by RT-PCR only in plants collected in Michoacán and Oaxaca (Ocampo et al., 2013). In wild poinsettia plants collected in Morelos, Guerrero and Nayarit, the presence of TMV was only detected by DAS-ELISA (^{Ocampo, 2012}).

Considering that Mexico imports commercial varieties of poinsettia for their production and commerce, that these varieties were developed from wild plants collected in Mexico, and that PnMV is one of the causes of economic losses in the production of this ornamental plant, the objective of this research was to search for a source of resistance to PnMV in wild poinsettia plants.

In 2013 and 2014, wild poinsettia plants with virus-associated symptoms and symptomless were collected in Jiutepec, Morelos (10 plants); Tehuilotepec, Guerrero (9 plants); Xilitla, San Luis Potosí (2 plants), and in a greenhouse at the Chapingo Autonomous University (UACh) in the State of Mexico (11 plants). Likewise, "Freedom" and "Red Prestige" commercial poinsettia plants were got at the "Vivero Internacional" company located in Morelos state. Shoots from two out of the nine plants collected in Tehuilotepec and two from Xilitla were placed into an isolated room in the virus greenhouse of the UACh Agricultural Parasitology Department.

The PnMV was detected using DAS-ELISA in young leaves and roots of all the plants collected in Jiutepec and Tehuilotepec, and only foliar tissue from the UACh greenhouse and commercial plants (5 "Freedom" and 4 "Red Prestige" plants) was analyzed. The detection protocol was carried out following the manufacturer instructions (Reagent Set SRA 90700/0096 Agdia^(r)). Absorbance values were measured after 60 min of incubation in an ELISA micro-plates reader (Multiskan EX, Labsystems) at 405 nm. Samples were considered positive when their absorbance values exceeded the average of the negative controls multiplied by 2 (^{García et al., 2014}).

RNA was isolated from 100 mg of leaf tissue of all the samples with the commercial Kit ZR Plant RNA MiniPrep(tm) (R2024 Zymo Research) by following the manufacturer protocol. In order to verify that there were no inhibitors in the ARN samples, one segment of the 18S gen was amplified using the specific primers 18SF (5´- ACG GAT CGC ACG GCC TTC GT -3) and 18SR (5´- ACC AGA CTT GCC CTC CAA TGG -3) (^{Zamboni et al., 2008}). A partial amplification of PnMV was done with PnMV-F (5´-GTG CCA GCC GCC GTT CTT CT-3´) and PnMV-R primers (5´-TGA GCC GGC GAC TCC ATC CA-3´) (^{Ocampo et al., 2013}). The positive control was gotten from poinsettia commercial plants of the "Red Prestige" variety, which were kept in the UACh virus greenhouse. The negative control was the reaction mix without the complementary ADN. PCR products were resolved in an 1.5 % w/v agarose gel at 100 V for 80 min. Amplicons were visualized with an UV transiluminator (Vilber Lourmat, Quantum), purified using the commercial Kit Wizard^(r) SV Gel and PCR Clean-Up System from Promega^(r) following the manufacturer protocol, and sent for sequencing to the UNAM Biotechnology Institute.

For transmission assays, PnMV was individually and mechanically inoculated in four wild poinsettia plants obtained from the stalks collected in Tehuilotepec and Xilitla, and from 10 species of indicator plants (Nicotiana benthamiana, N. clevelandii, N. tabacum cv. Xanthi, N. glutinosa, N. occidentalis, N. virginia, N. rustica, Datura stramonium, Chenopodium amaranticolor and Ch. quinoa). The inoculum source was obtained from the "Red Prestige" and "Freedom" commercial varieties that showed the highest virus concentration, according to the serological results. Briefly, three to five young leaves were sprinkled with 600-mesh carborundum and rubbed individually with a sterile cotton swab previously dipped in the inoculum. For a negative control, carborundum was applied on the surface of the leaves, then they were rubbed with a sterile cotton swab dipped into a pH 7.4 phosphate + diethyldithiocarbamate acid (DIECA) buffer.

Furthermore, young shoots of "Red Prestige" and "Freedom" plants that were PnMV positives by DAS-ELISA were grafted on three wild poinsettia plants obtained from stalks collected in Tehuilotepec, which were PnMV negative according to DAS-ELISA test. Each plant was cut obliquely; grafts were cut like a wedge shaped and inserted into the cuts made on the wild poinsettia plants, according to ^{Jayasinghe and Chuquillanqui (1992)} methodology. Plants were kept under observation in the UACh virus greenhouse. Two wild poinsettia plants were used as negative control controls, one as a graft (young shoots) and another as a rootstock (plants developed from stalks collected in Guerrero and San Luis Potosí).

Only some of the wild poinsettia plants collected in Jiutepec and in the UACh greenhouse showed chlorosis, leaf deformation, mosaic and ring spots symptoms (Figure 1). Mottling and leave deformation matched with symptoms induced by PnMV on poinsettia commercial plants grown in commercial nurseries in Venezuela (^{Carballo et al., 2001}) a New Zealand (^{Lebas et al., 2007}); however, the serological and molecular tests done in this research did not detect the virus. The 32 wild poinsettia plants were negative for the antiserum assessed. In general, the average absorbance of the leaf samples was 0.113 (Table 1); while root tissue assessed from the same plants collected in Tehuilotepec and Jiutepec, as well as one of the wild poinsettia plant from UACh green house, was 0.134 (Table 2). These results did not either match with those reported by ^{Ocampo et al. (2013)} in "nochebuena de sol" plants collected in Michoacán and Oaxaca, where the ELISA test showed that only 4 out of 28 plants with chlorosis, mosaic, variegation, leaf deformation and white spots symptoms were positive for PnMV. These symptoms, expressed after the plants were established at the INIFAP facilities (Zacatepec Experiment Field) from the twigs collected in the indicated sites, do not match with those induced by PnMV in commercial poinsettia plants, and they could be caused by insects. However, the fact that PnMV had been detected in the sap suggests that these samples became contaminated with the positive controls or that the infection caused by this virus is latent. If the infections are latent, it is possible that the symptomless wild poinsettia plants collected in Tehuilotepec, Jiutepec, and those developed from the twigs collected in Xilitla may provide a source of resistance to PnMV. However, as it was mentioned previously, in all these plants the DAS-ELISA test did not detect the virus, and the specific primers only amplified the expected band in the positive controls ("Freedom"). The 18S ribosomal primers retrotranscribed the expected RNA fragment that was isolated from the 32 wild poinsettia plants, showing a 99 % of similarity with the sequence recorded in the GenBank (EU326598.1) (data not shown). This suggests that the quality of the nucleic acid was good and that there were no reverse transcription inhibitors.

Figure 1 Leaves of wild poinsettia plants (Euphorbia pulcherrima) showing possible viral symptoms. Plant leaves kept in Chapingo Autonomous University's greenhouse. Mosaic (A y B), ringspots (C) and mottling (D). Leaves of plants collected in Jiutepec, Morelos. Deformation (E) and chlorotic spots (F). Asymptomatic leaves of plants collected in Tehuilotepec, Guerrero (G) and in Xilitla, San Luis Potosí (H). 

Table 1 Absorbance at 405 nm using DAS-ELISA for leaf tissue of wild poinsettia plants (Euphorbia pulcherrima). 

Detection limit (L. D.) = Negative control average x 2 = 0.1105 x 2 = 0.221.

Table 2 Absorbance at 405 nm using DAS-ELISA for root tissue of wild poinsettia plants (Euphorbia pulcherrima). 

Detection limit (L. D.) = Negative control average x 2 = 0.1105 x 2 = 0.221.

In improved poinsettia plants, PnMV cannot induce symptoms or it is able to cause foliar mottled, and leaf and bract distortion (^{Carballo et al., 2001}; ^{Lebas et al., 2007}). The "Freedom" and "Red Prestige" poinsettia used in this research showed mosaic and were positive for PnMV according to ELISA (Table 3) and RT-PCR results. In order to determine if the wild poinsettia plants were or not hosts for the virus, the foliar sap from "Freedom" plants that where PnMV positives by DAS-ELISA was mechanically inoculated on wild poinsettia leaves, but no symptoms were evident and the virus was not detected with the used techniques, nor it was defined if the grafts from the "Freedom" and "Red Prestige" varieties transmitted PnMV to the wild poinsettia plants because of graft abortion.

Table 3 Absorbance at 405 nm using DAS-ELISA for leaf tissue of two poinsettia commercial varieties (Euphorbia pulcherrima). 

Detection limit (L. D.) = Negative control average x 2 = 0.1105 x 2 = 0.221.

It is possible that the luminosity and temperature conditions during the transmission assays, that the biological features of the donor and recipient plants (phenological stage of development, nutrition) and/or that the characteristics of the buffer used to prepare the inoculum (associated with the virus stability and pathogenicity) have inhibited or prevented the mechanical and graft transmission of PnMV to wild poinsettias plants. However, it is also likely that the wild poinsettia are non-host plants for the virus, so they could be considered as a source of resistance to further development of Mexican varieties that can be integrated into an improvement genetic system to enhance their commercial features.

There was no transmission of PnMV to the indicator plants evaluated. N. benthamiana, which is one of the plants that has been infected with this virus (^{Lebas et al., 2007}), showed small necrotic spots (Figure 2E) that rather than viral symptoms look like mechanical damages caused during inoculation; however, control treatment leaves did not develop these symptoms. N. rustica, N. occidentalis and N. virginia leaves developed chlorotic spots (Figure 2A, ^C , ^G , ^I ), but further serological and molecular tests did not detect the virus.

Figure 2 Leaves of N. rustica (A y B), N. occidentales (C and D), N. benthamiana (E and F), N. tabacum cv. Xanthi (G y H) and N. virginia (I and J) 34 days after inoculation with PnMV. All plants developed chlorotic mottling, except for N. benthamiana, which showed small necrotic spots. B, D, F, H and J correspond to the negative control leaves. 

In contrast with the results from wild poinsettia plants, the commercial plants were positive for PnMV, according to RT-PCR and DAS-ELISA results. The partial sequence of the virus was 92 % similar to the AB550791.1 sequence reported by GenBank (data not shown); and the plants had an average absorbance of 0.610 ("Freedom") and 0.286 ("Red Prestige"), which were higher than the detection limit value (0.221) (Table 3). As previously stated, commercial poinsettia plants infected with this virus can be symptomless or develop chlorosis, and leaf and bract distortion (^{Carballo et al., 2001}; ^{Lebas et al., 2007}). PnMV is a worldwide-spread pathogen. It has been reported in Australia (^{Spetz et al., 2008}), Korea (^{Chung et al. 2004}), Denmark (^{Bech, 1996}), New Zealand (^{Lebas et al., 2007}), Venezuela (^{Carballo et al., 2001}), among other countries. In Mexico, this is the first time that the virus is reported in the poinsettia commercial plants that were evaluated.