Genome-Wide Identification, Expression, and Response to Fusarium Infection of the SWEET Gene Family in Garlic (Allium sativum L.)

Proteins of the SWEET (Sugar Will Eventually be Exported Transporters) family play an important role in plant development, adaptation, and stress response by functioning as transmembrane uniporters of soluble sugars. However, the information on the SWEET family in the plants of the Allium genus, which includes many crop species, is lacking. In this study, we performed a genome-wide analysis of garlic (Allium sativum L.) and identified 27 genes putatively encoding clade I–IV SWEET proteins. The promoters of the A. sativum (As) SWEET genes contained hormone- and stress-sensitive elements associated with plant response to phytopathogens. AsSWEET genes had distinct expression patterns in garlic organs. The expression levels and dynamics of clade III AsSWEET3, AsSWEET9, and AsSWEET11 genes significantly differed between Fusarium-resistant and -susceptible garlic cultivars subjected to F. proliferatum infection, suggesting the role of these genes in the garlic defense against the pathogen. Our results provide insights into the role of SWEET sugar uniporters in A. sativum and may be useful for breeding Fusarium-resistant Allium cultivars.


Introduction
Carbohydrates are important biomolecules for energy generation and storage in plants and, as such, are involved in plant development and stress response [1]. Furthermore, sugars define the economic value of many crops. Thus, the content and ratio of fructose and glucose in sink organs significantly affect the taste and quality of grape [2] and tomato [3]) fruit, potato tubers [4], and carrot roots [1]).
Glucose, a basic monosaccharide produced by photosynthesis, is used for the synthesis of fructose and sucrose [5]; then, the three sugars are transported from the leaves through the phloem to accumulating organs (flowers, fruit, seeds, and roots), where they are stored (partly as starch) to support organ growth and development [6,7]. To date, three families of sugar transporters have been characterized: monosaccharide transporters (MSTs), sucrose transporters (SUTs), and Sugar Will Eventually be Exported Transporters (SWEETs) [6,8].
Despite their extensive expansion and functional diversity, the SWEET genes are considered to be evolutionarily conserved. A study of this gene family in 31 representative plant species demonstrated a dramatic increase in the number of SWEET members in higher plants, largely due to tandem and segmental duplication under purifying selection [9]. Structurally, SWEET transporters are unique due to the presence of up to seven transmembrane domains, which may have arisen during the evolutionary duplication of a prokaryotic ancestral gene. In the plant, SWEETs are involved in maintaining a variety of physiological processes such as pollen nutrition, nectar secretion, seed filling, phloem loading, and pathogen feeding [10]. the same time, clade II SWEET genes are suggested to play a role in promoting pathogen virulence (AtSWEET4, AtSWEET5, AtSWEET7, and AtSWEET8 in Arabidopsis [16]) or be involved in plant response to necrotrophs (SWEET4 in Arabidopsis and grape [39][40][41]).
Garlic (Allium sativum L.) is the second most popular spice and the second most important crop in the Allium genus [42,43], with production of over 30 million tons per year (http://www.fao.org/; accessed on 26 January 2023). Garlic is known to have healthpromoting and anti-fungal properties, producing compounds that can regulate glucose metabolism and transport and exert therapeutic effects in cerebral ischemia and obesity [44,45] and volatile and non-volatile molecules that act as fungicides [46]. Garlic cloves are rich in carbohydrates, including high-molecular weight fructans and fructooligosaccharides [47,48], as well as soluble sugars such as sucrose, fructose, and glucose [49,50], which constitute up to 75% of the bulb dry matter. Most sugars are photoassimilate-derived, suggesting the existence of an extensive carbohydrate transportation system; however, the information on sugar transporters in garlic is lacking. In particular, nothing is known about SWEETs, not only in garlic, but also in other Allium species.
The aim of the present study was to search for and characterize the A. sativum genes belonging to the SWEET family and investigate their tissue expression patterns. As garlic is susceptible to Fusarium infection known to affect SWEET expression, we also examined SWEET gene activity in response to F. proliferatum attack in garlic cultivars resistant and susceptible to Fusarium basal rot (FBR).

Identification of SWEET Genes in the A. sativum Genome
Twenty-seven sequences of full-length SWEET genes were found by in silico analysis of the A. sativum cv. Ershuizao genome (PRJNA606385) and transcriptome (PRJNA607255) were annotated as A. sativum (As) SWEET1-27 (Table 1). Given the low homology between the genes of garlic and the model species A. thaliana, combined with the lack of data on SWEET genes in Allium species, the gene numbering was based on the order of their location on the chromosomes, as was done, for example, for maize Zea mays SWEETs [51].

Characterization of Putative SWEET Proteins in Garlic
The AsSWEET proteins predicted based on the CDSs consisted of 217-333 amino acids ( Table 1). Analysis of AsSWEET physicochemical properties indicated that the proteins differed in isoelectric points (pIs) (from 5.84 in AsSWEET16 to 10.01 in AsSWEET21), hydrophobicity index (GRAVY) (from 0.203 in AsSWEET14 to 0.860 in AsSWEET12), and molecular weight (MW) (from 23.991 kDa in AsSWEET13 to 34.077 kDa in AsSWEET23) ( Table 1).
Next, we analyzed the relationships of the predicted AsSWEETs and the SWEETs of A. thaliana, S. lycopersicum, and Z. mays since Arabidopsis, tomato, and maize are considered models for studying dicot and monocot species, respectively (1) and there are no Thus, besides three tandemly organized AsSWEET gene clusters on chromosomes 2, 8, and 7 (Figure 1a), we could also distinguish five phylogenetic clusters (AsSWEET26, AsS-WEET27; AsSWEET14, AsSWEET12, AsSWEET10, AsSWEET13; AsSWEET1, AsSWEET15, AsSWEET17; AsSWEET18, AsSWEET19, AsSWEET2; and AsSWEET6, AsSWEET3, AsS-WEET11) (Figure 3), the members of which were located on different chromosomes and might have been originated as a result of segmental/transposed gene duplication events.
Next, we analyzed the relationships of the predicted AsSWEETs and the SWEETs of A. thaliana, S. lycopersicum, and Z. mays since Arabidopsis, tomato, and maize are considered models for studying dicot and monocot species, respectively (1) and there are no data on SWEET genes in monocot Allium species for comparison (2). The results indicated that AsSWEETs of groups 5/6, 4, 1, and 2/3 ( Figure 3a) corresponded to A. thaliana SWEET clades I, II, III, and IV, respectively, and that the tandemly organized AsSWEET gene clusters ( Figure 1a) corresponded to specific functional A. thaliana SWEET clades: AsSWEET4-9, AsSWEET18, and AsSWEET19, to clade III and AsSWEET20-25, to clade IV ( Figure 4).

Figure 4.
Phylogenetic relationships of SWEET proteins from A. sativum (red) and A. thaliana (black). Analysis was performed using the neighbor-joining method in MEGA7.0.26 [53]. Percentages of replicate trees in which the associated sequences clustered together in the bootstrap test (1000 replicates) are shown next to the branches. Clades I-IV refer to SWEET clades previously identified in A. thaliana [16]. The numbering of maize genes is given in accordance with [51].
It was found that, within clades I, II, and IV, the garlic genes split into subclades based on similarity to genes of Arabidopsis, tomato (dicot), and maize (monocot). In clade I, AsSWEET1 was grouped with At/SlSWEET3 and ZmSWETT4/ZmSWETT16, Analysis was performed using the neighbor-joining method in MEGA7.0.26 [53]. Percentages of replicate trees in which the associated sequences clustered together in the bootstrap test (1000 replicates) are shown next to the branches. Clades I-IV refer to SWEET clades previously identified in A. thaliana [16]. The numbering of maize genes is given in accordance with [51].
Clade III had the largest number of members, but the SWEETs of garlic, as well as maize, were grouped predominantly separately from the SWEET proteins of dicotyledonous plant species. All AsSWEETs (AsSWEET4-9, AsSWEET18, and AsSWEET19) are included in the Z. mays ZmSWEET3, ZmSWEET6, ZmSWEET9, ZmSWEET19, and ZmSWEET320 subclade ( Figure 4). Similar clustering of clade III proteins, preferentially separating monocot and dicot proteins, is shown, for example, in [54].
In addition, NCBI-BLAST was used to determine the closest homologs of AsSWEET proteins in two model plant species, A. thaliana and S. lycopersicum (Table S1). It was found that the proteins of garlic have only 49-67% identity with the SWEETs of tomato and Arabidopsis. Clade III AsSWEETs were the most similar to AtSWEET11 and AtSWEET12 (except for AsSWEET16, which was close to AtSWEET14) in A. thaliana and to SlSWEET10c, SlSWEET12b, SlSWEET11c in S. lycopersicum. Clade I proteins were the closest homologs of AtSWEET3/SlSWEET3 (AsSWEET1), AtSWEET1/SlSWEET1b (AsSWEET15 and As-SWEET17) and AtSWEET2/SlSWEET2a-2 (AsSWEET26 and AsSWEET27). The highest similarity of clade II and IV proteins was observed with AtSWEET7/SlSWEET7a and AtSWEET17/SlSWEET16, respectively (Supplementary Table S1).
In total, 10 types of cis-elements related to stresses such as anaerobic conditions, drought, cold, wounding, and pathogens were identified in AsSWEET promoters, all of which contained at least one ( Figure 5). Two genes, AsSWEET4 and AsSWEET5, contained only stress-related elements; the former, along with AsSWEET8 and AsSWEET9, had the largest number (4) of elements responsive to anaerobic conditions. The largest number of STREs was found in AsSWEET8 (5), followed by AsSWEET23 (3). AsSWEET16 and AsSWEET17 contained only wound-responsive elements-two Wun-motifs and one W-box, respectively; the other wounding/pathogen-responsive motifs (WRE3 and/or box S) were found only in AsSWEET26 and 27. A drought-responsive element was identified only in AsSWEET10 ( Figure 5).  Nine types of hormone-responsive cis-elements, including those involved in the response to abscisic acid (ABA), auxin (AuxRR), methyl JA (MeJA), salicylic acid (SA), gibberellic acid (GA), and ethylene (ET), were identified in AsSWEET genes ( Figure 5). Each AsSWEET promoter had 2-5 elements, except for AsSWEET4 and 5, which had none. The Figure 5. Hormone-and stress-related cis-acting elements found in the promoter regions of AsSWEET genes. MeJA, methyl jasmonate. The color scheme (from pale to dark) corresponds to the numbers of cis-elements (from low to high).
The expression of 11 genes (AsSWEET1, AsSWEET15, AsSWEET17, and AsSWEET26 of clade I, AsSWEET13 of clade II, AsSWEET3, AsSWEET5, AsSWEET9, AsSWEET11, and AsSWEET18/19 of clade III, and AsSWEET24 of clade IV) was verified by quantitative real-time (qRT)-PCR in the tissues of A. sativum cv. Sarmat (Figure 7). It was confirmed that AsSWEET1 and AsSWEET13 were not transcribed in any garlic tissue. The results also revealed that the expression of most analyzed genes in the roots, cloves, leaves, and pseudostems was similar in cv. Ershuizao and Sarmat (Figures 6 and 7); however, three genes showed different expression levels: AsSWEET5 in the pseudostem (high/traces, respectively), AsSWEET15 in the roots (traces/high, respectively), and AsSWEET11 and AsSWEET24 in the leaves and pseudostems (high/low, respectively). "AsSWEET18-19" means the sum expression of AsSWEET18 and AsSWEET19 genes; the high identity (over 98%) of their mRNAs forced the design of common primers. The data were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and ubiquitin (UBQ) mRNA levels; a-h p < 0.01 indicates significant differences between tissues; ps.stems, pseudostems.
AsSWEET3, AsSWEET9, AsSWEET15, AsSWEET17, and AsSWEET26 were expressed in all analyzed tissues with the maximal level in the roots (AsSWEET26), leaves (AsSWEET5 and AsSWEET15), peduncles (AsSWEET3, AsSWEET9, AsSWEET11, AsSWEET17, and AsSWEET24), and receptacles (AsSWEET9 and AsSWEET18-19). AsSWEET5 was expressed only in the leaves. The expression of AsSWEET9 and AsSWEET17 was characteristic for the cloves and air bulbs, whereas that of AsSWEET5 and AsSWEET11 was absent in these organs, and that of AsSWEET5, AsSWEET18-19, and Figure 7. Transcription of the selected AsSWEET genes in A. sativum cv. Sarmat tissues. "AsSWEET18-19" means the sum expression of AsSWEET18 and AsSWEET19 genes; the high identity (over 98%) of their mRNAs forced the design of common primers. The data were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and ubiquitin (UBQ) mRNA levels; a-h p < 0.01 indicates significant differences between tissues; ps.stems, pseudostems.
AsSWEET3, AsSWEET9, AsSWEET15, AsSWEET17, and AsSWEET26 were expressed in all analyzed tissues with the maximal level in the roots (AsSWEET26), leaves (AsSWEET5 and AsSWEET15), peduncles (AsSWEET3, AsSWEET9, AsSWEET11, AsSWEET17, and AsSWEET24), and receptacles (AsSWEET9 and AsSWEET18-19). AsSWEET5 was expressed only in the leaves. The expression of AsSWEET9 and AsSWEET17 was characteristic for the cloves and air bulbs, whereas that of AsSWEET5 and AsSWEET11 was absent in these organs, and that of AsSWEET5, AsSWEET18-19, and AsSWEET24 was absent in the roots (Figure 7).
The expression of AsSWEET5, AsSWEET18, AsSWEET19, and AsSWEET24 was not detected either in non-infected or infected roots, which is consistent with the previous results (Figure 7), whereas that of the other six genes (AsSWEET3, AsSWEET9, AsSWEET11, AsS-WEET15, AsSWEET17, and AsSWEET26) was affected by F. proliferatum infection (Figure 8). At 24 hpi, AsSWEET3, AsSWEET11, AsSWEET15, and AsSWEET26 were significantly upregulated in both cultivars, especially in cv. Strelets, where AsSWEET9 was induced about 8-fold, AsSWEET3-over 20-fold, and AsSWEET11, AsSWEET15, and AsSWEET26-about 30-fold compared to the uninfected control ( Figure 8). Thus, the AsSWEET genes were strongly upregulated in the FBR-susceptible cultivar at the early stage of Fusarium infection. Later (96 hpi), the expression level of all AsSWEET genes in cv. Strelets decreased but still remained significantly higher than in control; at the same time, in cv. Sarmat, the expression of AsSWEET3 and AsSWEET11 was increased 140-150 times (Figure 8). These results indicated cultivar-dependent activation of AsSWEET genes in response to Fusarium infection.

Effect of F. proliferatum Infection on Sucrose, Glucose, and Fructose Contents in Garlic
Next, we compared the content of soluble sugars (glucose, fructose, and sucrose) in infected and non-infected garlic cultivars. The results indicated that at 24 hpi, the levels of all sugars were significantly decreased in the roots of FBR-resistant cv. Sarmat but not in those of FBR-susceptible cv. Strelets where the levels of glucose and sucrose were unchanged and that of fructose significantly increased (Figure 9). No significant changes in the sugar content were observed in both cultivars at 96 hpi. This may indicate that the FBR-resistant cv. Sarmat does not supply Fusarium with sugars, while the FBR-sensitive cv. Strelets supplies pathogen in an early response to infection but may stop efflux at 96 hpi. in those of FBR-susceptible cv. Strelets where the levels of glucose and sucrose were unchanged and that of fructose significantly increased (Figure 9). No significant changes in the sugar content were observed in both cultivars at 96 hpi. This may indicate that the FBR-resistant cv. Sarmat does not supply Fusarium with sugars, while the FBR-sensitive cv. Strelets supplies pathogen in an early response to infection but may stop efflux at 96 hpi.

Cloning and Characterization of CDSs and Regulatory Regions of AsSWEET Genes Differentially Expressed in FBR-Sensitive and -Resistant Cultivars after F. proliferatum Infection
Considering the differential expression of AsSWEET3, AsSWEET9, AsSWEET11, and AsSWEET26 genes in cv. Sarmat and Strelets in response to F. proliferatum infection (Figure 8), we cloned and sequenced the CDSs and regulatory regions of these genes and compared them with those of cv. Ershuizao (used as reference).

Cloning and Characterization of CDSs and Regulatory Regions of AsSWEET Genes Differentially Expressed in FBR-Sensitive and -Resistant Cultivars after F. proliferatum Infection
Considering the differential expression of AsSWEET3, AsSWEET9, AsSWEET11, and AsSWEET26 genes in cv. Sarmat and Strelets in response to F. proliferatum infection (Figure 8), we cloned and sequenced the CDSs and regulatory regions of these genes and compared them with those of cv. Ershuizao (used as reference).
Analysis of the regulatory regions (~1 kb, including promoter;~0.7 kb in case of AsSWEET11) indicated that those of AsSWEET3 had a 340 bp deletion and 3 SNPs in both cv. Sarmat and Strelets; in addition, this gene had a 3 bp insertion in cv. Sarmat and carried 2 SNPs in cv. Strelets. In both cv. Sarmat and Strelets, the promoter regions of AsSWEET9 had a 2 bp deletion and 12 SNPs, those of AsSWEET11-2 SNPs, and those of AsSWEET26-8 SNPs. In addition, the AsSWEET26 promoter of cv. Sarmat carried four extra SNPs and a 10 bp deletion, whereas that of cv. Strelets had 10 extra SNPs ( Table 2).

Discussion
SWEET uniporters that mediate the bidirectional transfer of soluble sugars between different tissues are important not only for plant growth and development but also for the immune response [13,28,29,31,32].
Accumulating evidence indicates that during infection, plant pathogens selectively target SWEET transporters to access host carbohydrate reserves and obtain soluble sugars [39]. Thus, Xanthomonas spp. are known to upregulate the expression of specific SWEET genes; conversely, the blockage of SWEET activation can increase plant resistance to infection and reduce pathogen growth and virulence [39]. It is suggested that the suppression of SWEET genes in plants can reduce their sensitivity to infection and may potentially be employed as a strategy to breed resistant crop cultivars [31,32,39]. However, this approach is not entirely correct since symbiotic arbuscular mycorrhizal (AM) fungi, which can significantly enhance plant growth and adaptability by colonizing their roots, also obtain carbohydrates from the root cortex apoplast, including through the SWEET activities. For example, colonization of potato roots by the AM fungus Rhizophagus irregularis causes significant changes in the transcription of 22 out of 35 genes of the SWEET family [59]. Therefore, functional studies of the specific SWEET genes can help to determine the balanced effect on the transporter activity, which would contribute to plant protection, on the one hand, and, on the other hand, would not break the symbiotic potential of plant-microbial communities.
The present study was the first attempt to identify and characterize SWEET genes in the Allium genus in general and in garlic in particular. Overall, we found 27 SWEET genes in A. sativum cv. Ershuizao through genome-wide analysis (Table 1). Multiple SWEET genes are characteristic of many species, including agricultural crops; thus, there are 17 such genes in A. thaliana, 25 in Musa acuminate Colla, 29 in S. lycopersicum, and 108 in Triticum aestivum L. [60]. However, if in tomato, for example, the size of gene families can be regulated through sexual reproduction and introgressive hybridization [60][61][62], it is not the case in garlic, which is an asexually reproducing (apomictic) species whose genotypes presumably originate through mutations in a single clone, leading to significant phenotypic diversity [63]. It can be speculated that the high number of SWEET genes in garlic is a result of accelerated random mutagenesis and somaclonal variation observed especially in the genes related to the basic processes of plant development and adaptability [60][61][62]. Alternatively, it is possible that the SWEET gene family in A. sativum had evolved before this species lost the ability for sexual reproduction.
The observed clustering of AsSWEET20-25 (clade IV) and AsSWEET18/19 and AsSWEET4-9 (clade III) on chromosomes 8, 7, and 2, respectively ( Figure 1a) suggests their evolutionary origin through tandem duplication and points on the functional redundancy of paralogous transporters within these groups. Differences in the expression patterns of AsSWEET genes in the three clusters, including complete silencing of some members (Figures 6 and 7), may reflect the neofunctionalization effect of duplicated genes due to mutations during adaptation to environmental changes. Such mutations can occur in the CDSs as well as in the regulatory sequences, as evidenced by the obvious differences in the patterns of cis-regulatory elements in the AsSWEET promoter regions ( Figure 5).
FBR of the bulbs and Fusarium wilt of the leaves can lead to significant (up to 60%) losses of the garlic crop [64][65][66][67][68][69]. F. proliferatum, one of the Fusarium species infecting garlic [57], is a common hemibiotrophic pathogen that utilizes plant sugars both in the biotrophic and necrotrophic phases [70]. Therefore, we hypothesized that this fungus could activate carbohydrate metabolism in garlic through the regulation of AsSWEET genes. Analysis of AsSWEET transcription in the roots of FBR-resistant and -susceptible garlic cultivars at the biotrophic (24 hpi) and pre-necrotrophic (96 hpi) phases of F. proliferatum infection revealed the upregulation of six genes, AsSWEET3, AsSWEET9, AsSWEET11, AsSWEET15, AsSWEET17, and AsSWEET26, in response to infection in both cultivars (Figure 8). Among these genes, AsSWEET3, AsSWEET9, and AsSWEET11 belong to clade III (Figure 4), the members of which are known to be induced in plants at the biotrophic phase of pathogen infection [39]. However, there was a significant difference in the activation level of AsSWEET3, AsSWEET9, and AsSWEET11 between FBR-resistant and -susceptible cultivars both at 24 and 96 hpi (Figure 8), which is consistent with the role of clade III SWEETs in response to Fusarium infection. Thus, the expression of these genes was much more significantly upregulated in the sensitive cv. Strelets than in the resistant cv. Sarmat at 24 hpi (Figure 8), suggesting pathogen-induced activation in order to access host sugars and promote colonization at the biotrophic stage. At the same time, AsSWEET3, AsSWEET9, and AsSWEET11 were much stronger upregulated in the resistant cv. Sarmat than in the susceptible cv. Strelets at the pre-necrotrophic (96 hpi) stage of infection ( Figure 8). It is possible that the late-stage induction of AsSWEET3, AsSWEET9, and AsSWEET11 in the resistant cultivar could represent an attempt of the host plant to stimulate sugar outflow from the infected roots in order to restrict pathogen growth, which is in agreement with a decrease in fructose, glucose and sucrose levels in the roots of cv. Sarmat but not in those of cv. Strelets (Figure 9), as well as the fact that sugars act as signaling molecules and interact with the hormonal network that regulates the immune system of plants [71]. Differential expression patterns of SWEET genes have also been reported in Fusarium-resistant and -susceptible cultivars of watermelon [36]. Thus, our results, together with previous data, indicate distinct roles of individual SWEETs in plant resistance to fungal infection.
The three genes of clade I (AsSWEET15, AsSWEET17, and AsSWEET26) upregulated by F. proliferatum infection showed overall similar dynamics and degree of activation in both cultivars (Figure 8), suggesting that the regulatory pathways underlying AsSWEET15, AsSWEET17, and AsSWEET26 transcriptional induction in response to infection are not involved in the garlic resistance to Fusarium.
Despite the differential transcriptional response to the infection of AsSWEET3, As-SWEET9, AsSWEET11, and AsSWEET26, sequence alignment revealed a high degree of structural similarity in their cis-regulatory regions and protein products in FBR-susceptible and -resistant garlic cultivars (Tables 2 and 3). This fact points to intricate molecular mechanisms controlling SWEET gene expression in garlic, which are likely genotype-dependent and defines the strength of garlic resistance to Fusarium infection.
Earlier studies of FBR-resistant and -susceptible garlic cultivars have identified a set of A. sativum PR genes encoding chitinases, glucanases, and thaumatin-like proteins, which also have infection stage-dependent activation patterns and could be involved in anti-fungal immune defense [55][56][57]. Together with these findings, our present results suggest that AsSWEET genes can be a part of a complex protection system evolved in garlic to resist biotic as well as abiotic stresses, which requires further investigation.

Identification and Structural Analysis of the A. sativum SWEET Genes
The search for the full-length SWEET genes was performed in the A. sativum cv. Ershuizao genome (PRJNA606385, assembly Garlic.V2.fa) according to the annotated transcriptome (PRJNA607255) (genome and transcriptome were prepared by Sun et al. (2020) [52]); only sequences with start and stop codons and two MtN3_slv domains (pfam03083) in the translated version were selected. Comparative gene and protein structural analyses were conducted with MEGA 7.0.26 [53]. The phylogenetic dendrogram was constructed based on protein sequences using MEGA 7.0.26 (Neighbor-Joining method); confidence for tree topologies was estimated by bootstrap values of 1000 replicates. Used in analysis A. thaliana and S. lycopersicum SWEET numbering is given according to [27] and [21,72], respectively.

In Silico mRNA Expression Analysis
AsSWEET gene expression in garlic tissues was analyzed based on A. sativum cv. Ershuizao RNA-seq dataset (ID: PRJNA607255), which was normalized as Fragments Per Kilobase of transcript per million mapped reads (FPKM) [52] and included data on the gene expression profiles in the roots, bulbs (8 developmental stages), leaves, pseudostems, buds, flowers, and sprouts. According to Sun et al. (2020) [52], buds were collected at the bud stage of garlic plants planted in October 2018; samples of roots, leaves, sprouts, pseudostems (212 days after sowing), and flowers (217 days after sowing) were collected in May 2019; samples of bulbs of eight growth stages were taken every five days from April 11 to May 16, 2019. Transcripts with an average FPKM value of ≥10 in at least one tissue type were selected and used for heatmap construction with Heatmapper [74].

Plant and Fungi Material and F. proliferatum Infection
FBR-resistant cv. Sarmat and FBR-susceptible cv. Strelets used in this study are winter garlic cultivars of Russian breeding. Bulbs from the 2022 harvest were kindly provided by the Federal Scientific Vegetable Center (Moscow region, Russia).
The F. proliferatum strain was originally isolated from the bulbs of garlic cv. Strelets was kindly provided by the Group of Experimental Mycology, Winogradsky Institute of Microbiology (Research Center of Biotechnology of the RAS, Moscow, Russia). According to the pathogenicity test, the first signs of the disease appear on the clove surface 5 days after infection [55].
F. proliferatum infection was performed as previously described [56]. Cloves were surface sterilized in 70% ethanol for 3 min, rinsed with sterile water, placed in Petri dishes with wet filter paper, and incubated at +25 • C in the dark for 72 h until active root growth was observed. Then, cloves were infected by soaking in F. proliferatum conidial suspension (~10 6 conidia mL −1 ) for 5 min, transferred to fresh Petri dishes, and incubated at +25 • C in the dark for 24 and 96 h (n = 3 cloves per each time point); uninfected cloves were used as control. The roots, pseudostems, and cloves were collected at each time point, frozen in liquid nitrogen, and stored at −80 • C.

RNA Extraction and qRT-PCR Analysis
Total RNA was extracted from garlic tissues (0.1 g of each) using the RNeasy Plant Mini Kit (QIAGEN, Hilden, Germany), cleaned from genomic DNA (RNase free DNase set; QIA-GEN), qualified by gel electrophoresis, and used for first-strand cDNA synthesis (GoScript Reverse Transcription System; Promega, Madison, WI, USA) with an oligo-dT primer. RNA and cDNA concentrations were quantified by fluorimetry (Qubit ® Fluorometer, Thermo Fisher Scientific, Waltham, MA, USA). qRT-PCR was performed in a CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA) with 3.0 ng cDNA, SYBR Green RT-PCR mixture (Syntol, Moscow, Russia), and specific primers. Primers were designed based on A. sativum cv. Ershuizao transcriptomic data (PRJNA607255) (Table  S2) as a result of comparative analysis of AsSWEET mRNAs; the most variable regions of each gene were used to select primers so that there was at least one intron between them. Further, manual revision of sequence polymorphisms and additional evaluation was performed using Primer3 (http://frodo.wi.mit.edu/primer3/, accessed on 15 October 2022). Primers were experimentally tested at three concentrations of cDNA (mix from various organs)-10, 1, and 0.1 ng per well in three technical replicates. Primer pairs with only one significant peak occurring in the melt curve above the instrument detection limit were selected; the efficiency of the validated pairs of primers was 95-110%, R 2 = 0.943-0.999. Given the high identity (over 98%) of the AsSWEET18 and AsSWEET19 mRNAs, common primers were designed for these genes (Table S2). The following cycling conditions were used: initial denaturation at 95 • C for 5 min and 40 cycles of denaturation at 95 • C for 15 s, and annealing/extension at 60 • C for 40 s. mRNA expression of AsSWEET genes was normalized to that of two reference genes, GAPDH and UBQ [56], which were previously used separately as a reference in onion (Allium cepa L.) and garlic [75,76]. We used the combination of GAPDH and UBQ based on studies showing that the combination of two stably expressed reference genes, rather than one, improves the accuracy of qPCR (for example, [77]); previously, we have already successfully applied this approach to the analysis of garlic gene expression [55][56][57]. The results were statistically analyzed with GraphPad Prism version 8 (GraphPad Software Inc., San Diego, CA, USA; https://www.graphpad.com/scientific-software/prism/ [accessed on 10 January 2023]). The data were expressed as the mean ± standard error (SE) based on three technical replicates of two biological replicates for each combination of cDNA and primer pairs. The unequal variance (Welch's) t-test was applied to assess differences in gene expression; p < 0.01 was considered to indicate statistical significance.

Gene Amplification and Sequencing
To amplify the AsSWEET CDSs, gene-specific primers were designed based on A. sativum cv. Ershuizao transcriptomic data (NCBI project accession number: PRJNA607255) (Supplementary Table S2). PCR amplification was performed with 30 ng cDNA from the roots of each cultivar at the following conditions: initial denaturation at 95 • C for 5 min, 35 cycles of denaturation at 95 • C for 30 s, annealing at 55 • C for 30 s, and extension at 72 • C for 2 min, and final extension at 72 • C for 5 min. PCR products of the expected size were purified by using the QIAEX ® II Gel Extraction kit (QIAGEN, Hilden, Germany), cloned in the pGEM ® -T Easy vector (Promega, Madison, WI, USA), and sequenced (3-5 clones for each accession) on ABI Prism 3730 DNA Sequencer (Applied Biosystems, Waltham, MA, USA) using the same primers.

Promoter Analysis
The search for cis-elements in the promoters (~1.0 kb regions upstream of the initiation codon) was performed using the PlantCARE database of cis-regulatory elements, enhancers, and repressors; (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/; accessed on 26 January 2023). Promoter sequences and defining of hormone-and stress-responsive cis-elements are provided in Supplementary Tables S3 and S4, respectively.

Conclusions
We identified and characterized 27 genes encoding sugar uniporters of the SWEET family in the A. sativum cv. Ershuizao genome and cloned AsSWEET3, AsSWEET9, AsS-WEET11, and AsSWEET26 CDSs and promoter regions from FBR-resistant and -susceptible garlic cultivars. Structural analysis of AsSWEET genes indicated conserved functional involvement in the bidirectional transport of soluble sugars: hexoses (clade I AsSWEET1, AsSWEET15, AsSWEET17, AsSWEET26, and AsSWEET27 and clade II AsSWEET10, and AsSWEET12-14), fructose (clade IV AsSWEET20-25), and sucrose (clade III AsSWEET2-9, AsSWEET11, AsSWEET16, AsSWEET18, and AsSWEET19). The AsSWEET promoters contained hormone-and stress-related elements associated with the response to fungal pathogens. Comparative AsSWEET transcriptional profiling in the F. proliferatum-infected roots of FBR-resistant and -susceptible cultivars suggests that clade III AsSWEET3, As-SWEET9, and AsSWEET11 genes could be involved in the garlic response to Fusarium infection. Our results provide the first insights into the functions of SWEET genes in the plants of the Allium genus and may be used in breeding programs to increase the resistance of Allium crops to Fusarium infections.   Table 2).