**3. Results**

To determine the location of the *Wx* gene, different combinations of primers were used. However, for some of them, owing to similar fragment sizes, establishing the unambiguous presence of the *Wx* gene from *H. chilense* was difficult. The BDFL/BRD primers, designed by Nakamura et al. [33], provided the most reliable results (Figure 1).

**Figure 1.** (**a**) Diagrammatic representation of *Wx* gene showing the three fragments used for sequencing, and (**b**) PCR analysis for the chromosomal location of *H. chilense Wx* gene using primers BDFL/BRD from Nakamura et al. [33] in common wheat, ditelosomic addition lines, and *H. chilense*. Lanes are as follows: 1, cv. Chinese Spring (CS), 2, CS + 7*HchS* line, 3, CS + 7*HchL* line, and 4, *H*1 line.

Because data obtained from other *Triticeae* species indicated that the *Wx* genes were mainly located on the short arm of the chromosome 7 [10], the CS + 7HchS and CS + 7HchL lines were used to determine the arm location of this gene in *H. chilense*. Figure 1 shows the presence of one additional band in the CS + 7*HchL* line (Lane 3) that is a similar size to a band in the *H. chilense* (H1) line (Lane 4), which is absent in both the common wheat CS (Lane 1) and the CS + 7*HchS* line (Lane 2). Thus, the *Wx-Hch*1 gene is located on 7*HchL* (H1), suggesting an inversion in this *H. chilense* chromosome.

The *Wx*-*Hch*1 gene was analyzed in two lines of *H. chilense* that represent two different biotypes of this species present in Chile. Due to the length of this gene (~2700 bp), the genomic sequence was obtained amplifying three fragments, which covered the complete coding sequence. The first fragment of ~620 bp, covered part of the second exon (the first one in the coding sequence, see Figure 1) until the end of the fourth exon, while the second fragment (~960 bp) spanned the fourth to the seventh exons. Finally, the third fragment (~1160 bp) was the region between the end of fragment 2 and the 12th exon, including the TGA codon. The alignment and comparison are shown in Figure S1. The initiation codon, ATG, and the termination codon, TGA, for translation, as well as the splice junctions of each intron of *Wx-Hch1*, were in homologous positions to those in other *Wx* genes. Both alleles detected in *H. chilense* (*Wx-Hch1a* and *Wx-Hch1b*) were smaller in size than the *Wx* genes used for comparison (Table 1).

The comparison between the seven nucleotide sequences showed that the greatest homology level was detected between the *Wx*-*Hch*1 genes and the *Wx*-*H*1 variants from barley (90.8%). However, the comparison of the *Wx* genes from common wheat cv. Chinese Spring showed lower values of 83.6% for *Wx*-*D*1, 84.7% for *Wx-B*1 and 87.7% for *Wx-A*1. Nevertheless, the predicted proteins of these same sequences showed homology greater than 94% for all comparisons, and greater than 97.7% among barley species. This is in concordance with most of the sequence differences being found in introns. In all cases, the exons were the same size, with the exception of exon 2, which contained one or two additional codons in wheat but was similar in both *H. chilense* and *H. vulgare* (Table 1).


**Table 1.** Size of the di fferent exons and introns of the coding sequence in the *Wx* sequences evaluated.

1 cv. Chinese Spring (NCBI ID: *Wx-A*1, AB019622, *Wx*-*B*1, AB019623, *Wx*-*D*1, AB019624) [43]. 2 cv. Vogelsanger Gold (NCBI ID: X07931) [12]. 3 cv. Morex (NCBI ID: AF474373).

### *3.1. Amino acid Predicted Sequence Analysis*

While coding sequences of the *Wx* genes varied, most variation resulted in silent mutations that did not impact protein sequence or structure. Additionally, these proteins were synthesized as precursors or pre-proteins, including one transit-peptide of 70 amino acids and one mature domain. Nevertheless, potentially impactful sequence variation was detected in a conserved region of the mature domain related to waxy protein activity. These changes can lead to marked di fferences in the predicted sequences of the respective proteins (Figure S2).

The *Hordeum* sequences, including both *H. chilense* biotypes, showed the insertion of one amino acid residue within the signal peptide between positions 53 and 54. Furthermore, these sequences had deletions of Gly73 or Ala73 residues detected in the wheat proteins. Both InDels were the consequences of the aforementioned elimination of one or two codons in exon 2. Three non-conservative amino acid changes were detected in both *H. chilense* variants. For two of these changes, Pro24 → Arg and Ser416 → Pro, the H7 line showed the same amino acids as the other evaluated sequences. The Ser416 → Pro change could have deleterious e ffects according to the PROVEAN analysis, with a score of −2.869. The 419 position was di fferent in both *H. chilense* sequences and was also di fferent than the other sequences, with the exception of Wx-A1, which was similar to that of Wx-Hch1a (Table 2).

The amino acid sequences from *H. chilense* were more similar to the waxy proteins from barley than those derived from any wheat genome. Only 15 changes were detected between *H. chilense* and barley variants, while up to 54 changes were observed when this comparison was carried out with wheat waxy proteins, and ten changes were common to both barley and wheat species (Figure S2).

Up to 15 amino acids variants were detected within the transit peptide. The sequence changes generated some variations in the secondary structure of the transit peptide. The most dramatic was the Val5 → Ala change (detected in barley waxy proteins but not in *H. chilense* ones) that resulted in an elongated first helix and the disappearance of a β strand in the secondary structure (Figure 2). With the exception of the aforementioned change in position 24 (Pro in Wx-Hch1a and Arg in the others), all changes were common to both *H. chilense* variants. The barley variants showed three of

these conservative changes (Val5 → Ala, Ile18 → Val and Met68 → Val), although the waxy protein of cv. Vogelsanger Gold (*Wx-H1a* variant) included one additional non-conservative change at position 70 (Arg70 → Ser). Four of these differences, all classified as conservative, were common to Wx-B1 and Wx-D1 (Pro25 → Ala, Leu30 → Val, Asn34 → Ser, and Ser62 → Thr), although each variant showed two additional changes (Ala39 → Pro and Ile45 → Thr for Wx-B1, and Ile45 → Val and Lys52 → Thr for Wx-D1). Nevertheless, Wx-A1 was the most varied, with five unique changes (three conservative: Ile18 → Val, Ala58 → Pro and Gly61 → Phe, and two non-conservative: Gly17 → Ser and Ser62 → Asp) and one change in common with Wx-D1 (Ile45 → Val) (Figure 2).


**Table 2.** Amino acid comparison of predicted mature protein among waxy protein variants evaluated.

1 This position should be increased for *Wx*-*A*1 and -D1 (+1), and for *Wx-B*1 (+2). 2 NCBI ID: barley [X07931] and common wheat cv. "Chinese Spring" [*Wx-A*1: AB019622, *Wx-B*1: AB019623, *Wx*-*D*1: AB019624].

**Figure 2.** Comparison of the secondary structure motifs predicted by Garnier for the transit peptide region among all sequences evaluated.

Numerous changes were also observed in the mature proteins both in barley and wheat (Table 2). Five of these changes were common to both species (Asn123 → Lys, Thr142 → Arg, Pro 232 → Leu, Ser416 → Pro and Asp590 → Glu), whereas two changes were exclusively detected in barley (Arg438 → Lys and Ala496 → Val) and nine were exclusively detected in wheat (Val115 → Ile, Phe145 → Tyr, Ile158 → Val, Trp162 → Cys, Ala373 → Gly, Ile427 → Val, Leu471 → Val, Ala535 → Val and Gln551 → His). The other changes were detected in one or two wheat sequences (Table 2). Two of these changes could have effects on enzyme function (Figure 3). In addition to the abovementioned change, Ser416 → Pro, which was unique to Wx-Hch1a, another change with deleterious effects predicted by the PROVEAN analysis was Pro232 → Leu, with a score of −4.061. The other changes observed were considered neutral.

**Figure 3.** Consensus sequence of the predicted proteins from *H. chilense* showing the motifs described by Leterrier et al. [44] conserved in waxy proteins. Squares indicate substitution sites (blue for wheat, red for barley and black for both ones), and arrows point relevant changes found in the novel alleles.

Up to five of the eight motifs described by Leterrier et al. [44] are involved in the ADP glucose-binding and catalytic sites within the mature protein. Three changes were observed inside these conserved motifs (Figure 3). However, only the change Ile427 → Val was considered relevant because, although this change was also detected in barley waxy protein, the wheat waxy proteins all showed Val as the residue in this position. The change Thr356 appeared in barley and the *Wx*-*D*1 protein, while Val 496 was exclusively found in the barley protein. The PROVEAN analysis suggested that these changes be considered neutral because their influence on the enzyme function was very limited.
