Genome-Wide Characterization of Nitrogenase Reductase (nifH) Genes in the Sweet Potato [Ipomoea batatas (L.) Lam] and Its Wild Ancestors

The sweet potato (Ipomoea batatas (L.) Lam.) is an important and widely grown crop, and the nitrogenase reductase (nifH) gene is the most widely sequenced marker gene used to identify nitrogen-fixing bacteria and archaea. There have been many examples of the isolation of the diazotrophic endophytes in sweet potatoes, and there has been no report on whether sweet potatoes and their wild ancestors harbored nifH genes. In this study, a comprehensive analysis of nifH genes has been conducted on these species by using bioinformatics and molecular biology methods. A total of 20, 19 and 17 nifH genes were identified for the first time in sweet potatoes, I. trifida and I. triloba, respectively. Based on a phylogenetic analysis, all of the nifH genes, except for g10233.t1, itf14g14040.t1 and itb14g15470.t1, were clustered into five independent clades: I, II, III, IV and V. The nifH genes clustered in the same phylogenetic branch showed a more similar distribution of conserved motifs and exons–introns than those of the other ones. All of the identified genes were further mapped on the 15 chromosomes of the sweet potato, I. trifida and I. triloba. No segmental duplication was detected in each genome of three Ipomoea species, and 0, 8 and 7 tandemly duplicated gene pairs were detected in the genome of the sweet potato, I. trifida and I. triloba, respectively. Synteny analysis between the three Ipomoea species revealed that there were 7, 7 and 8 syntenic gene pairs of nifH genes detected between the sweet potato and I. trifida, between the sweet potato and I. triloba and between I. trifida and I. triloba, respectively. All of the duplicated and syntenic nifH genes were subjected to purifying selection inside duplicated genomic elements during speciation, except for the tandemly duplicated gene pair itf11g07340.t2_itf11g07340.t3, which was subjected to positive selection. Different expression profiles were detected in the sweet potato, I. trifida and I. triloba. According to the above results, four nifH genes of the sweet potato (g950, g16683, g27094 and g33987) were selected for quantitative real-time polymerase chain reaction (qRT-PCR) analysis in two sweet potato cultivars (Eshu 15 and Long 9) under nitrogen deficiency (N0) and normal (N1) conditions. All of them were upregulated in the N1 treatment and were consistent with the analysis of the RNA-seq data. We hope that these results will provide new insights into the nifH genes in the sweet potato and its wild ancestors and will contribute to the molecular breeding of sweet potatoes in the future.


Introduction
The sweet potato, Ipomoea batatas (L.) Lam., is an important food crop widely grown in the world. It is also an alternative source of bioenergy as a raw material for fuel production [1,2]. The sweet potato contains soluble sugar, starch, dietary fiber, protein, fat, calcium and other minerals, as well as antioxidant substances beneficial to human health, such as carotenoids, anthocyanins, vitamins, flavonoids, etc. [2]. Therefore, the sweet potato is rated as the healthiest vegetable by the World Health Organization. predicted gene model annotation, were downloaded from the Ipomoea Genome Hub (https://ipomoea-genome.org/, accessed on 18 August 2021); the genome sequences of I. trifida and I. triloba, including the predicted gene model annotation, were downloaded from GenBank BioProject (accessions numbers PRJNA428214 and PRJNA428241). Both a BLAST search and a hidden Markov model search (HMMsearch) were performed as described previously [31]. All protein sequences were first searched for the Fer4_NifH domain (Pfam accession number: PF00142) using hmmsearch with default parameters. Moreover, the extended amino acid sequence of Fer4_NifH domain was used as a query to search for all protein sequences in the sweet potato genome using the BLASTP program. After that, the genes gained by HMMsearch and BLAST methods were merged, and the redundant ones were removed.

Conserved Motif Detection and Gene Structure Analyses of the nifH Genes
To investigate the structural motif diversity of the identified nifH genes, the protein sequences of them were subjected to motif analysis by online MEME SUITE (https:// meme-suite.org/meme/, accessed on 19 July 2022) [39]. The criteria used for MEME analysis were (1) a minimum width of 6; (2) a maximum width of 50; (3) a maximum number of motifs designed to identify 20 motifs; and (4) iterative cycles set by default. The exon-intron structure of the nifH genes was acquired from the GFF3 annotation files of the sweet potato, I. trifida and I. triloba. The distribution of conserved motifs and the exonintron structures of the nifH genes were exhibited using TBtools software (version 1.068) (https://github.com/CJ-Chen/TBtools/releases, accessed on 16 July 2021) [40].

Duplication and Ka/Ks Analysis of the nifH Genes
To search for potential duplicated nifH genes in the sweet potato, I. trifida, and I. triloba, the Multiple Collinearity Scan toolkit (MCScanX, version 0.8) (http://chibba.pgml.uga.edu/ mcscan2/, accessed on 3 January 2020) was used [42]. All the nifH protein sequences of the three species were compared to themselves by using the BLASTP program with an E-value of 1 × 10 −10 . The resulting blast hits were incorporated along with chromosome coordinates of all nifH genes as an input for MCScanX analysis. The hits were classified into various types of duplications, including segmental, tandem, proximal and dispersed under a default criterion. The final results were drawn by CIRCOS software for visualization [43]. The aligned protein sequences of the nifH genes of the sweet potato, I.trifida and I.triloba were first converted into the corresponding nucleotide sequences using PAL2NAL software (http: //www.bork.embl.de/pal2nal/#RunP2N, accessed on 25 July 2021) [44] and were then summited to PAML software (version 4.0) (http://abacus.gene.ucl.ac.uk/software/paml. html, accessed on 11 July 2020) [45] for Ka/Ks (nonsynonymous/synonymous) calculation.

Expression Profiles of nifH Genes of the Sweet Potato, I. trifida and I. triloba
For the expression profile analysis of the nifH Genes in the sweet potato, I. trifida and I. triloba, RNA-Seq datasets were downloaded from the sequence read archive (SRA) of NCBI, which referred to different tissues of sweet potatoes (PRJNA511028), and the expressional information (fragments per kilobase of exon model per million mapped fragments, FPKM) of I. trifida and I. triloba was acquired from the sweet potato Genomics Resource (http://sweetpotato.uga.edu/gt4sp_download.shtml, accessed on 13 December 2021). After removing the low-quality reads and adaptor trimming, the clean RNA-Seq reads were aligned to the genome sequences of the sweet potato via Hisat2 [46]. Thereafter, SAMtools software (version 1.11) was used for aligned read counting (https://github. com/samtools/samtools/releases/download/1.11, accessed on 23 December 2020) [47]. Then, the obtained read counts were imported into DEseq2 for the analysis of differentially expressed genes (DEGs) [46]. For each compared course, it was treated as a DEG if |log2FC| > 1 and FDR ≤ 5%, and a mean log2FC value for each gene was calculated. The heat map was produced to distribute the expression levels using the RPKM (i.e., reads per kilobase per million) value in MeV software (version 4.9.0) (https://sourceforge.net/ projects/mev/, accessed on 23 August 2020) [48].

RNA Isolation and Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR) Analysis
Two sweet potato cultivars (Eshu 15, medium long vine; Long 9, short vine) were selected for the qRT-PCR analysis of the nifH genes. Freshly cut sweet potato seedlings, with 20-to 30-cm-long stems and five to seven leaves, were dipped into water for transplanting for 3 days. Then, the seedlings were transferred into a Hoagland nutrient solution, and the nutrient solution was set with the following nitrogen levels: N0 (0 mmol·L −1 pure nitrogen) and N1 (14 mmol·L −1 pure nitrogen). Other components remained the same: , and the first-strand cDNA was prepared using EasyScript All-in-One First-Strand cDNA Synthesis SuperMix for qPCR (One-Step gDNA Removal) (TransGen, Wuhan, China). The sweet potato β-actin gene (Genbank AY905538) was selected and used to normalize the relative quantities of the target genes. Three replications were performed, and the expression changes were calculated using the 2 -∆∆Ct method for each sample. Then, a quantitative real-time polymerase chain reaction (qRT-PCR) for four sweet potato nifH genes (g950, g16683, g27094 and g33987) was performed. The primers used for PCR were designed using on line Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/, accessed on 21 July 2022) (Table S1).

Identification of the nifH Genes of the Sweet Potato, I. trifida and I. triloba
A total of 20, 19 and 17 nifH genes were identified in the genomes of the sweet potato, I. trifida and I. triloba, respectively (Table 1). Among the 20 nifH genes in the sweet potato, the shortest (g56705.t1) was of 67 amino acids, whereas the longest (g4650.t1) was of 939 amino acids, and the average length of the genes was 430.55 amino acids (Table 1). Among the 19 nifH genes in I. trifida, the shortest (itf12g22980.t1) was of 289 amino acids, whereas the longest (itf12g19260.t1) was of 530 amino acids, and the average length of the genes was 372.21 amino acids (Table 1). Among the 17 nifH genes in I. triloba, the shortest (itb11g03920.t3) was of 278 amino acids, whereas the longest (itb14g15470.t1) was of 611 amino acids, and the average length of the genes was 385 amino acids (Table 1).

Phylogenetic Analysis of the nifH Genes of the Sweet Potato, I. trifida and I. triloba
To analyze the phylogenetic relationship of the nifH genes in the sweet potato, I. trifida and I. triloba, a phylogenetic tree was constructed ( Figure 1). All of the nifH genes, except for g10233.t1, itf14g14040.t1 and itb14g15470.t1, were clustered into five independent clades: I, II, III, IV and V, with support values > 93% ( Figure 1). Each of the five clades contained the nifH genes from all of the three species, which indicated that the ancestries for each clade were differentiated before the species specification of the nifH genes. Among the genes that were not clustered, the same phenomenon was detected; g10233.t1, itf14g14040.t1 and itb14g15470.t1 were identified from the genome of the sweet potato, I. trifida and I. triloba, respectively ( Figure 1). Although the number of the nifH genes in the sweet potato, I. trifida and I. triloba for each clade was different, it meant that the nifH genes of each clade may have experienced different duplication during the species specification of the nifH genes ( Figure 1).

Conserved Motif Detection and Gene Structure Analyses of the nifH Genes
Conserved motifs were detected in all of the identified nifH genes except for g25326.t1 in the three Ipomoea species (Figure 2). Among these conserved motifs, motif 1 was the most conserved one, and 54 of the 56 nifH genes harbored motif 1, followed by motif 2 (39 of 56), motif 3(27 of 56) and motif 6 (25 of 56). Of the identified nifH genes, itb14g15470.t1 contained the greatest number of conserved motifs (#14), followed by itb11g07630.t1 (#10). It was also found that the nifH genes clustered in the same phylogenetic branch showed a more similar distribution of conserved motifs and exonsintrons (gene structure) than those of the other nifH genes (Figure 2).

Chromosome Locations of the nifH Genes of the Sweet Potato, I. trifida and I. triloba
Based on the locations of individual nifH genes, all of the identified genes were mapped on the 15 chromosomes of the sweet potato, I. trifida and I. triloba, respectively ( Figure 3). However, the chromosome distribution of the nifH genes was similar between I. trifida and I. triloba, whereas it was different between the sweet potato and I. trifida or I. triloba (Figure 3). The nifH genes in the sweet potato were distributed on all of the chromosomes except for chromosome 4, 11 and 13, whereas the nifH genes in I. trifida and I. triloba were mainly located on chromosome 5, 6, 9, 11, 12 and 14 ( Figure 3).  Based on the locations of individual nifH genes, all of the identified genes were mapped on the 15 chromosomes of the sweet potato, I. trifida and I. triloba, respectively ( Figure 3). However, the chromosome distribution of the nifH genes was similar between I. trifida and I. triloba, whereas it was different between the sweet potato and I. trifida or I. triloba (Figure 3). The nifH genes in the sweet potato were distributed on all of the chromosomes except for chromosome 4, 11 and 13, whereas the nifH genes in I. trifida and I. triloba were mainly located on chromosome 5, 6, 9, 11, 12 and 14 ( Figure 3).

Duplication and Ka/Ks Analysis of the nifH Genes
The duplication analysis of the nifH genes in the three Ipomoea species showed that no segmental duplication was detected in each genome of them, and 0, 8 and 7 tandemly duplicated gene pairs were detected in the genome of the sweet potato, I. trifida and I. triloba, respectively ( Figure 3). The synteny analysis between the three Ipomoea species revealed that there were 7, 7 and 8 syntenic gene pairs of nifH genes detected between the sweet potato and I. trifida, between the sweet potato and I. triloba and between I. trifida and I. triloba, respectively (Figure 4).

Duplication and Ka/Ks Analysis of the nifH Genes
The duplication analysis of the nifH genes in the three Ipomoea species showed that no segmental duplication was detected in each genome of them, and 0, 8 and 7 tandemly duplicated gene pairs were detected in the genome of the sweet potato, I. trifida and I. triloba, respectively ( Figure 3). The synteny analysis between the three Ipomoea species revealed that there were 7, 7 and 8 syntenic gene pairs of nifH genes detected between the sweet potato and I. trifida, between the sweet potato and I. triloba and between I. trifida and I. triloba, respectively (Figure 4).   The non-synonymous substitution (Ka) to synonymous substitution (Ks) ratio (Ka/Ks) is an informative value of positive selection. To detect whether some nifH genes are under positive selection, Ka/Ks analysis was performed on duplicated and syntenic nifH genes within or between the studied three Ipomoea species. All of the duplicated and syntenic gene pairs showed a Ka/Ks ratio <1, except for the tandemly duplicated gene pair itf11g07340.t2_itf11g07340.t3, which has a Ka/Ks ratio = 1.34 ( Table 2). The results revealed that itf11g07340.t2_itf11g07340.t3 was subjected to positive selection, and all of the other duplicated and syntenic nifH genes were subjected to purifying selection inside duplicated genomic elements during speciation.

Expression Patterns of the nifH Genes in the Sweet Potato, I. trifida and I. triloba
In the sweet potato, nearly half of the identified nifH genes (9 genes) were nonregulated in all of detected tissues. The others were mainly upregulated in the fibrous roots (FR) and leaves and were mainly downregulated in the proximal ends (PE) and root stalks (RS) ( Figure 5). In the I. trifida, 8, 6, 1, 4, 4 and 2 of the 19 identified nifH genes were upregulated in the callus flowers, callus stems, flowers, flower buds, leaves and stems, respectively, and other genes in the tissue mentioned above or the 19 identified nifH genes in other tissues (i.e., root1 and root2) were mainly downregulated ( Figure 5). In the I. triloba, 3, 1, 10, 5 and 4 of the 17 identified nifH genes were upregulated in the flowers, flower buds, leaves, root1, root2 and stems, respectively, whereas other genes in the tissue mentioned above were mainly downregulated ( Figure 5).

qRT-PCR Analysis of the nifH Genes under Treatments
According to the above results, four nifH genes of the sweet potato (g950, g16683, g27094 and g33987) were selected for qRT-PCR analysis. All of the four genes were upregulated in N1 treatment and were consistent with the above analysis of RNA-seq data ( Figure 6). Compared with the control conditions (0 h), the transcripts of the four nifH genes all peaked after 3 h N1 treatment in the two sweet potato culitivars. Then, the expression level declined, and the expression level of Eshu 15 was higher than that of Long 9 ( Figure 6). In Eshu 15, g950, peaked at 3 h with an 8.11 fold higher expression level than that of the control, which was of the highest expression levels of the four nifH genes, followed by g27094 (7.99 fold), g33987 (3.66 fold) and g16683 (3.58 fold). In Long 9, g27094, peaked at 3 h with a 6.60 fold higher expression level than that of the control, which was of the highest expression levels of the four nifH genes, followed by g950 (5.16 fold), g33987 (3.21 fold) and g16683 (2.44 fold) ( Figure 6). However, in the N0 treatment, the transcripts of g950, g16683 and g33987 had all almost no change compared with the control conditions (0 h), and the transcripts of g27094 peaked after 3 h N0 treatment, with expression levels of only 1.69 and 1.69 fold higher than that of the control in Eshu 15 and Long 9, respectively ( Figure 6).

qRT-PCR Analysis of the nifH Genes under Treatments
According to the above results, four nifH genes of the sweet potato (g950, g16683, g27094 and g33987) were selected for qRT-PCR analysis. All of the four genes were upregulated in N1 treatment and were consistent with the above analysis of RNA-seq data ( Figure 6). Compared with the control conditions (0 h), the transcripts of the four nifH genes all peaked after 3 h N1 treatment in the two sweet potato culitivars. Then, the expression level declined, and the expression level of Eshu 15 was higher than that of Long 9 ( Figure 6). In Eshu 15, g950, peaked at 3 h with an 8.11 fold higher expression level than that of the control, which was of the highest expression levels of the four nifH genes, followed by g27094 (7.99 fold), g33987 (3.66 fold) and g16683 (3.58 fold). In Long 9, g27094, peaked at 3 h with a 6.60 fold higher expression level than that of the control, which was of the highest expression levels of the four nifH genes, followed by g950 (5.16 fold), g33987 (3.21 fold) and g16683 (2.44 fold) ( Figure 6). However, in the N0 treatment, the transcripts of g950, g16683 and g33987 had all almost no change compared with the control conditions (0 h), and the transcripts of g27094 peaked after 3 h N0 treatment, with expression levels of only 1.69 and 1.69 fold higher than that of the control in Eshu 15 and Long 9, respectively ( Figure 6). Figure 6. Expression analysis of g950, g16683, g27094 and g33987 of Eshu 15 and Long 9. The significance of differential gene expression levels compared with control are denoted as * < 0.05, ** < 0.01.

Discussion
Nitrogen is an essential element for plants, and it is also one of the most important limiting factors for obtaining a high agricultural yield. Biological nitrogen fixation is an important part of the terrestrial nitrogen cycle, which contributes 90-130 Tg N to the biosphere every year and is mainly completed by bacteria and leguminous plants [49]. nifH is a marker gene, and researchers have been able to characterize aspects of the diversity and ecology of nitrogen-fixing bacteria and archaea. The biological nitrogen fixation of plants, rather than by association with microorganisms, can generate crops that are less dependent on synthetic nitrogen fertilizers and can increase agricultural productivity and sustainability [50]. The sweet potato, the seventh largest food crop in the world [2], grows well in nitrogen-poor infertile soils, and it is believed that sweet potato endo-and epiphytic microorganisms play important roles in acting upon this growth ability in this plant [21][22][23][24]. Are these all the reasons? Do the sweet potato and its wild ancestors harbor nifH genes, even with nitrogen fixation potentiality?
Previous studies have demonstrated that, among the 291 tested accessions of cultivated sweet potatoes, all contained one or more transfer DNA (T-DNA) sequences, which is believed to be harbored by Agrobacterium [51]. The study suggested that an agrobacterium infection occurred in evolutionary times, and the T-DNA integration, the interruption of an F-box gene and the subsequent fixation of foreign T-DNA into the sweet potato genome occurred during the evolution and domestication of the sweet potato [51]. Based on the above demonstration, the sweet potato and its wild ancestors harboring nifH genes is possible.
In order to answer the question mentioned above, in this present study, the genome-wide identification of nifH genes was conducted for the first time in this study. A total of 20, 19 and 17 nifH genes were identified in the genome of the sweet potato, I. trifida and I. triloba, respectively. The number of nifH genes in the three investigated species was comparable. The phylogenetic analysis revealed that the identified nifH genes can be clustered into five independent clades: I, II, III, IV and V, with high support values. Moreover, each of the five clades contained the nifH genes from all of the three species, which indicated that the ancestries for each clade were differentiated before the

Discussion
Nitrogen is an essential element for plants, and it is also one of the most important limiting factors for obtaining a high agricultural yield. Biological nitrogen fixation is an important part of the terrestrial nitrogen cycle, which contributes 90-130 Tg N to the biosphere every year and is mainly completed by bacteria and leguminous plants [49]. nifH is a marker gene, and researchers have been able to characterize aspects of the diversity and ecology of nitrogen-fixing bacteria and archaea. The biological nitrogen fixation of plants, rather than by association with microorganisms, can generate crops that are less dependent on synthetic nitrogen fertilizers and can increase agricultural productivity and sustainability [50]. The sweet potato, the seventh largest food crop in the world [2], grows well in nitrogen-poor infertile soils, and it is believed that sweet potato endo-and epiphytic microorganisms play important roles in acting upon this growth ability in this plant [21][22][23][24]. Are these all the reasons? Do the sweet potato and its wild ancestors harbor nifH genes, even with nitrogen fixation potentiality?
Previous studies have demonstrated that, among the 291 tested accessions of cultivated sweet potatoes, all contained one or more transfer DNA (T-DNA) sequences, which is believed to be harbored by Agrobacterium [51]. The study suggested that an agrobacterium infection occurred in evolutionary times, and the T-DNA integration, the interruption of an F-box gene and the subsequent fixation of foreign T-DNA into the sweet potato genome occurred during the evolution and domestication of the sweet potato [51]. Based on the above demonstration, the sweet potato and its wild ancestors harboring nifH genes is possible.
In order to answer the question mentioned above, in this present study, the genomewide identification of nifH genes was conducted for the first time in this study. A total of 20, 19 and 17 nifH genes were identified in the genome of the sweet potato, I. trifida and I. triloba, respectively. The number of nifH genes in the three investigated species was comparable. The phylogenetic analysis revealed that the identified nifH genes can be clustered into five independent clades: I, II, III, IV and V, with high support values. Moreover, each of the five clades contained the nifH genes from all of the three species, which indicated that the ancestries for each clade were differentiated before the species specification of the nifH genes. All of the nifH genes could be located on chromosomes of the three species, and the distribution of them on sweet potatoes were different from I.trifida and I.triloba. Previous reports have suggested that the whole-genome triplication (WGT) occurred in an ancient ancestor of the Ipomoea lineage around 46.1 million years ago (Mya), much earlier than the divergence of I. nil from the lineage containing I. trifida and I. triloba (~3.6 Mya) and the I. trifida-I. triloba divergence (~2.2 Mya). The results of the comparison between the genomes of I. trifida (or I. triloba) and I. nil limited large-scale interchromosomal rearrangements over the last 3.6 million years [52], and in the sweet potato, two recent whole-genome duplication (WGD) events occurred about 0.8 and 0.5 million years ago [53]. Therefore, the two recent whole-genome duplications in the sweet potato may be the reason for the discrepancies of chromosome distribution.
Conserved motif detection and gene structure analyses showed that the detected motifs behaved with different degrees of conservation among the nifH genes, and the nifH genes clustered in the same phylogenetic branch showed a more similar distribution of conserved motifs and exons-introns than that of the other nifH genes. Similar results have been found in other gene families in various species, such as the WRKY gene family in pineapples [54], the AP2/ERF gene family in buckwheat [55], the superoxide dismutase (SOD) gene family in rapeseed [56], etc.
Segmental and tandem duplications have significantly contributed to gene family expansion in plants [57,58]. The duplication analysis in the present study shows that no segmental duplication was found in each genome of the three Ipomoea species, and 0, 8 and 7 tandemly duplicated gene pairs were detected in the genome of the sweet potato, I. trifida and I. triloba, respectively. The results suggest that there were no segmental duplications of nifH genes through polyploidy followed by chromosome rearrangements, and in several members of them in I. trifida and I. triloba, there occurred duplications within the same intergenic region or in neighboring intergenic regions of their genomes [59]. A total of 7, 7 and 8 syntenic gene pairs of nifH genes were detected between the sweet potato and I. trifida, between the sweet potato and I. triloba and between I. trifida and I. triloba, respectively. Similar results have been demonstrated by other closely related species. For example, in the Brassicaceae family, a number of NBS loci were identified: Arabidopsis lyrata (#78), A. thaliana (#58), Brassica rapa (#100), Capsella rubella (#52) and Thellungiella salsuginea (#59) [57]. These may be generated by different extents of the gene duplication of the ancestor genes [60]. The Ka/Ks analysis of the duplicated and syntenic nifH genes in the three Ipomoea species revealed that nearly all of the duplicated and syntenic nifH genes were subjected to purifying selection inside duplicated genomic elements during speciation, since they had a ratio of Ka/Ks < 1 [61].
The expression profile of the nifH genes showed that all but nine of the sweet potato genes detected upregulation or downregulation in various tissues of their corresponding species. Four nifH genes of the sweet potato were selected for qRT-PCR analysis, and the expressions were consistent with the transcriptome data analysis. It was also found that the nifH genes expressed differently in the three species, and the nifH genes expressed differently in the different tissues of the same species as well. The results suggest that the nifH genes experienced functional differentiation after the whole genome duplication. The phenomenon of the member of the family gene acting differently in expression profiles has been reported in various studies, for example, in wheat [62], tomatoes [63], tea [64], etc., and even in the duplicated genes [65]. In potatoes, it was found that there was variation within tandemly duplicated genes among cultivated, non-cultivated and wild potato genotypes in terms of bias in functional specificities, the proportion of lineage-specific clusters, diverged expression and promoter similarities [66].

Conclusions
In this study, A total of 20, 19 and 17 nifH genes were identified for the first time in the sweet potato, I.trifida and I.triloba, respectively. Following the identification, a phylogenetic tree was formed to cluster the identified nifH genes into five independent clades. All of the nifH genes could be located on chromosomes of the three species, and the distribution of them on sweet potatoes were different from I.trifida and I.triloba. The expression profiles revealed that the nifH genes were expressed differently in various tissues of the three species. A total of 4 nifH genes of the sweet potato were selected for qRT-PCR analysis in two sweet potato cultivars (Eshu 15 and Long 9) under nitrogen deficiency (N0) and normal (N1) conditions. The results acquired in this study may provide new insight into the nifH genes in the sweet potato and its wild ancestors, and they may contribute to the nitrogen-fixing breeding of sweet potatoes in the future.