Plant material and source of viruses

A total of 151 symptomatic plant samples were collected, including cultivated Capsicum chinense Habanero-type (45) and Scotch Bonet-type (12), C. annuum var. annuum landraces regionally identify as Dulce (18), X´kat iik (12), and Ya´ax iik (8) plants of C. annumm. var. aviculare locally called Maax iik (43), and C. frutescens (8) Tabasco-type plants that grow wild or in-home gardens. Additionally, we collected representative weeds from different families: Callicarpa spp., (Verbenaceae) (1), Croton malvaviscifolius (Euphorbiaceae) (1), Dicliptera sexangularis (Acanthaceae) (1), Herissantia crispa (Malvaceae) (1), and Nicotiana tabacum (Solanaceae) (1). Plant material was collected in locations within the states of Yucatan (YS) and Campeche (CS) in the YPM (Table 1, Figure 1). The weed leaf samples were gathered from wild grown plants in non-cropping sites and from pepper cultivated areas during a severe whitefly infestation seasons (February-May and July to the beginning of October of 2017). The samples were stored -20 ºC in the molecular virology laboratory of CICY, to serve as a source of viral DNA. Total genomic DNA from the leaves of pepper and weed plants was extracted according to the methods of ^{Dellaporta et al. (1983}) and ^{Echevarria-Machado et al. (2005}), respectively. Leaf samples were cleaned with sterile water before the DNA extraction in order to eliminate the possible contamination of whiteflies.

Figure 1 Map indicating the locations in the YPM, where symptomatic plants were collected for this study. Symbols represent the locations were the symptomatic plants were collected. 

Polymerase chain reaction (PCR) amplification, cloning, sequencing, and sequence analysis

The primer pair prAV324/AC889 was used for PCR amplification of the begomoviruses diagnostic ~576 bp core coat protein (core Cp) gene fragment (^{Brown et al., 2005}), following the protocol recommended by the authors. Positive virus-specific PCR products of ~576 bp were purified with a QIAEX II kit (QIAGEN) and cloned into pGEM T-Easy vector (Promega, Madison, WI, USA) following the manufacturer’s instructions. The nucleotide sequences were determined with an ABI 377 sequencer (Davis Sequencing Co., CA, USA). Most clones were sequenced in forward and reverse directions. To enhance the possibility of detecting multiple begomoviral genotypes due to mixed infections in field samples, three to five PCR products were cloned and their nucleotide sequences analyzed. The nucleotide sequences (120) from randomly selected cloned samples (52) were edited manually to remove degeneracy imposed by the primers (EditSeq, DNASTAR version 5.08, Madison, WI, USA), resulting in a 533 nt sequence of the core Cp, and partial identification of Begomovirus species was done by comparing with sequences available in the NCBI-GenBank database (http://www.ncbi.nlm.nih.gov) by BLASTn (^{Altschul et al., 1997}).

Sequence data analysis

For genetic and phylogenetic analysis, data of 25 selected core Cp partial begomoviruses sequences identified in the present study were used, 23 of which were isolated from species and landraces of Capsicum and two from different weed species. In addition, 36 reference standard sequences were obtained by cropping the core Cp of the full-length begomoviral genome from the GenBank database (Table 2), and included in the phylogenetic analysis. Multiple sequence alignment of the taxa was prepared with the CLUSTAL W in MegAlign (DNASTAR version 5.08, Madison, WI). Genetic distance matrix, diversity and pattern of nucleotide substitutions (total, synonymous and non-synonymous) analyses were complete with 23 PepGMV CP selected nucleotide sequences (Table 2). Selected PepGMV sequences included three samples of each Capsicum species and one sequence of each weed plant; the CP genomic region (CPR) comprised position 147-679 according to GenBank sequence AF077025. Molecular evolutionary testing was conducted using the Kimura 2-parameter and Kumar methods, Bootstrap analysis with 1000 replicates, with MEGA version 11 (^{Tamura et al., 2021}). Codon-based Z-test analysis averaging over all sequence pairs was performed to test for the selective force operating on CP sequences, using the Kumar method in MEGA 11. All positions containing gaps and missing data were eliminated (complete deletion option). There were a total of 177 positions in the final dataset. To reconstruct a phylogenetic tree, multiple aligned sequences were analysed by multiple parsimony using MEGA 11 with 1000 iterations per search. The strict consensus tree was constructed using the standard settings of MEGA 11, and statistical support for each major clade was calculated using Bootstrap with 1000 iterations per search. The tree was rooted using the CP sequence edited from the core CP sequence of the Maize streak virus (MSV-CP) of the genus Mastrevirus, family Geminiviridae (Table 2).

Search for viral symptoms in cultivated and field plants

In a disease survey during high whitefly infestation seasons at locations of the YS and CS (Figure 1), virus-like symptoms were collected on diverse cultivated and non-cultivated pepper and weed plants. All of sampled cultivated pepper plants, namely C. chinense, C. a. var. annuum, landrace Dulce, X´kat iik and Ya´ax iik, C. frutescens, and wild grown C. a. var. aviculare Maax iik, showed symptoms of golden mosaics, distorted leaves and stunting growth. Symptoms, clearly distinguished in C. frutescens, were often moderately perceived in one or two branches of Maax iik, rarely affecting the whole plant (Figure 2A). On the other hand, landrace X´kat iik (Figure 2B), Chile Dulce (Figure 2C and D) and Habanero chili (Figure 2E and F) plants showed noticeable foliar symptoms of varied severity, including yellowing with downward curly leaves, golden mosaics, and stunting growth. Comparable symptoms were recorded for weed species (images not shown). Nevertheless, in the majority of cases, chili plants showing small curled leaves were stunted and bushy in appearance because of reduced length internodes (Figure 2D).

Figure 2 Symptoms observed in wild-grown (A) and cultivated (B, C, D, E and F) plants sampled at Catmis (A), Muna (B), Merida (C), Cuzama (D), and Tekax (E and F) infected with Begomovirus (PCR+). (A) Capsicum annuum var. aviculare (Maax iik) plant showing branches with golden mosaic and distorted leaves, and stunting growth; (B) C. annuum var. annuum (X´kat iik) with severe golden mosaic, and distorted leaves; (C and D) C. annuum var. annuum (Chile Dulce) with severe golden mosaics, distorted leaves, and stunting growth; (E and F) C. chinense (Habanero type) with golden mosaics, and with downward curly leaves and severe stunting growth. 

Begomoviruses preliminary identification

Of the total collected plant samples (151), including 146 peppers and 5 weeds, 90.1% were PCR positive using the core CP primer (Table 3). That is, 89.7% of the peppers (131), including plant samples of all the pepper species and all the weed (5) plant samples were positive to begomoviruses. Sequencing of amplicons, on average two to three per sample of plants (52), uncovered the likely incidence of single and mixed begomovirus infections. Of these, virus species determined in 120 sequenced clones were provisionally identified based on the comparison of nucleotide identity (>89% nt) with viral core CP sequences of well-studied begomoviruses (Table 3). Single infections with Pepper golden mosaic virus PepGMV predominated, representing 82.5% of the sequenced clones (99) detected in materials of the different Capsicum species and in the weed species Callicarpa spp., C. malvaviscifolius, D. sexangularis, H. crispa, and N. tabacum (Table 3). This was followed by Pepper huasteco yellow vein virus PHYVV (5 %) found in two C. chinense plants, one clone isolated from Habanero and the other from Scotch Bonnet type, plus the Tomato severe leaf curl virus ToSLCV (4.2%) found in the Habanero and Dulce plants. Interestingly, four sequenced clones (3.3%) from a single Habanero chili plant (sample H6 EuMV-YP[C. chinense H]) displayed high identity (98-99%) to the Euphorbia mosaic virus-[Yucatan Peninsula] EuMV-YP isolate described from Euphorbia heterophylla (^{Hernandez-Zepeda et al., 2007b}). Possible mixed infections were confirmed by bi-directionally searching of three to five clones of each partial begomoviral species isolated from each plant. These included a mixture of PepGMV-PHYVV in the Habanero chili (sample 120, C. chinense H), the PepGMV-TYLCV in the Dulce pepper (sample 138, C. annuum D) and the PepGMV-ToSLCV in a X´kat iik plant (sample 144, C. annuum, Table 3).

Genetic variation of PepGMV isolates based on nucleotide distance analysis from partial DNA-A begomovirus sequences

Genetic distance analysis based on pairwise analysis of 23 chosen core CP nucleotide sequences (Table 4) indicated that PepGMV isolates were closely related strains sharing from 95 to 99% nucleotide sequence identity; percentage nt identity was calculated as [1-genetic distance] x100 (^{Brown et al., 2005}). The average nt distance of CPR estimated for the number of base substitutions per site over all 23 sequence pairs was equal to 0.027 (±0.004 SE), with a calculated transitional to transversional distance per site ratio (R) equal to 0.641(± 0.21 SE). In general, the estimated average codon-based divergence of all sequence pairs showed a low value of nonsynonymous (dN) substitutions to synonymous (dS) substitutions, dN = 0.003 < dS = 0.130, and a low dN / dS ratio of 0.023 (Table 4), respectively.

Phylogenetic analysis based on partial DNA-A begomovirus sequences

The phylogenetic analysis examined the probable relationship between begomovirus-infected pepper and those of weed species collected in the YS and CS relative to their corresponding sequences from a database (Gen-Bank). The analysis incorporated the nt sequences of the Coat protein gene (CP) core region from 61 taxa. The matrix was constructed with 23 nt sequences isolated from Capsicum spp., two from plants grown as weeds (Table 3), and 36 from Gen-Bank sequences (Table 2). The evolutionary history was inferred using the Maximum Parsimony method. The most parsimonious tree with length =3191 was show (Figure 3). The consistency index is 0.376378 (0.309704), the retention index is 0.634911 (0.634911) and the composite index is 0.238966 (0.196634) for all sites and parsimony-informative sites (in parentheses). The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) were show next to the branches. The MP tree was obtained using the Subtree-Pruning-Regrafting (SPR) algorithm with search level 1 in which the initial trees were obtained by the random addition of sequences (10 replicates). There were a total of 566 positions in the final dataset. Evolutionary analyses were conducted in MEGA11. The cluster of Eastern Hemisphere begomoviruses with a 99% Bootstrap supported value includes the Tomato yellow leaf curl virus, of which clone 138-TYLCV-[C. annuum D], isolated from a single plant with a likely mix infection (Table 3, Figure 3), showed high identity (99%) to the Dominican Republic isolate of TYLCV. Interestingly, a divergent clade containing three clones of PepGMV isolates from different C. annuum cultivars was observed close to the MSV-CP, the Mastrevirus that was selected as root of the phylogenetic tree, with 89% Bootstrap supported value. The Western Hemisphere cluster divides into two subclades (63% support node). One group holds the Tomato severe leaf curl virus clone 144-ToSLCV-[C.annuum X]), isolated from a X´kaat ik plant with a probable mixed infection (Table 3, Figure 3). This is related to the ToSLCV-[GT] reference isolate (82% Bootstrap), shares common nodes with other begomoviruses, and is grouped with species of Sida golden mosaic virus (SiGMV), Tomato mottle virus (ToMoV), Chino del tomate virus (CdTV), and Abutilon mosaic virus (AbMV). The second group, with two branches, contains the pepper isolate of the Euphorbia mosaic virus isolate, H6 EuMV-YP (C. chinense H) with its sister species, the EuMV-YP reference isolate (99% Bootstrap) and the Tomato mild yellow leaf curl Aragua virus (TMYLCAV), a relation supported by a 77% value (Figure 3). The former groups are assembled with a begomovirus species that received no Bootstrap support, including the SLCV clade. The other relationships inferred for the collections with no Bootstrap support are subdivided into two subclades. One subclade comprising the Pepper huasteco yellow vein virus (PHYVV) cluster is grouped with three isolates from C. chinense, clones 120 and 183 of PHYVV- from Habanero and the 40-PHYVV isolate from the Maax iik plant, with 88 to 99 % Bootstrap values. The second subclade is composed of collections of the Pepper golden mosaic virus PepGMV from the most of the pepper and weed isolates set with well-known begomovirus reference isolates forming a major PepGMV clade supported by Bootstrap (83%). This PepGMV clade visibly split into three subclades (Bootstrap 89%) (Figure 3). Two clones of PepGMV, R3 and 164 were grouped in a separate clade with TYLCV-DR, with 89%% of Bootstrap support. A future analysis with the complete sequence of the A Component of both clones should be necessary, in order to corroborate this phylogenetic relationship.

Figure 3 Cp sequence-inferred phylogenetic relationships of begomoviruses identified in field-grown and cultivated pepper plants from Yucatan and Campeche, Mexico; the analysis also includes the Cp sequence of selected, well-studied begomoviruses which are available in GenBank. Presented is the single most parsimonious tree; numbers placed at nodes represent Bootstrap values over than 50%, lower values are not shown (1000 iterations).