The reference genome (Arachis hypogaea v1.0) and adjoining information (protein sequence, mRNA, coding sequence, and DNA) were obtained from the Phytozome database (Goodstein et al., 2012). The ALDH domain (accession number PF00171) was downloaded from the PFAM database. The ALDH members were identified through hmmsearch and hmmbuild using the perl script in the Linux system, in which 1e⁻¹⁵ was set as a filter threshold (Zhang et al., 2020). The SMART software (Letunic and Bork, 2018) was used to confirm the ALDH domain and remove duplicates.

The evolutionary relationship of ALDH members was analyzed using neighbor-joining methods and a Poisson model implemented in MEGA X, with 1000 bootstrap repetitions (Kumar et al., 2018). The motif of ALDH members was analyzed using the MEME platform, in which the length of motifs was 10–15 amino acids, while the e-value of motifs was less than e⁻⁵ (Bailey et al., 2009). The gene structure of the ALDH members was analyzed using the GSDS software (Hu et al., 2015). The cis-acting elements were identified and predicted using PlantCare, in which the function of each cis-acting element was predicted (Lescot et al., 2002). The expression of ALDH members and their location on chromosomes was extracted from the Phytozome database (Goodstein et al., 2012). The results were plotted using TBtool (Chen et al., 2020). The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases were used for the gene annotation of ALDH members (Chen et al., 2017); Majorbio Cloud provided the platform for enrichment analysis (Shanghai, China).

The locally grown groundnut cultivar Silihong was used in this study. The seeds were provided by the Institute of Economic Botany, Heilongjiang Academy of Agricultural Sciences (Harbin, Hei Longjiang, China). Seeds of the same size were sterilized with 5% NaClO solution for 15 min, rinsed with distilled water, and placed in a Petri dish lined with double-layer filter paper. The seeds were covered with a single filter paper. Three replicates of 10 seeds per Petri dish were prepared for saline-alkali and control treatments. Petri dishes were placed in a GXZ intelligent light incubator (Ningbo Jiangnan Instrument Factory, Zehjiang, China) at 25°C in the dark. When the length of the peanut buds reached half the length of the peanut seeds, the seeds were sampled and used for expression pattern analysis. Control seeds were treated with distilled water. The rest of the seeds were subjected to saline-alkali treatment by adding saline-alkali solution (pH, 8.9) to Petri dishes. The sprouts in each Petri dish were set as the experimental unit, while each treatment had three experimental units as biological replicates.

The radicle was used as samples in this study, with biological and technological replicates. ALDH enzyme activity was determined under water and saline-alkali treatment at 0, 12, 24, 48, 72, and 96 h using a Multiskan FC microplate reader (Thermo Scientific Company, Wilmington, DE, USA) and the Elisa Kit (M0608, Michy Biology, Suzhou, China) according to the manufacturer’s instructions. Radicles were used as samples. The samples at 0 and 48 h from both treatments were used for quantitative real-time PCR (qRT-PCR) analysis. Total RNA was extracted using a MolPure^® Plant Plus RNA Kit (19292ES, Yeason, Shanghai, China). The qualified RNA was used for reverse transcription using a kit (11139ES, Yeason) after detection using a Nanodrop OneC (Thermo Fisher Scientific, Waltham, MA, USA) and 1% agarose for RNA quality. The primers for ALDH amplification were designed using DNAMAN (Lynnon Biosoft, San Ramon, CA, USA) (Nong et al., 2019), and the UKN1 gene was used as the reference gene ( Table S2 ). qRT-PCR was performed on the Roche platform (480 II, Roche, Basel, Switzerland) using Hieff UNICON^® Universal Blue qPCR SYBR Green Master Mix (11184ES, Yeason) according to the manufacturer’s instructions. The expression levels of ALDH members were calculated using the 2^−ΔCt method (Livak and Schmittgen, 2001). Data were analyzed in the SPSS software (Bezzaouha et al., 2020). Duncan’s multiple method was used to test differences between groups.

One hundred and five ALDH members were identified from the reference genome (Arachis hypogaea v1.0) from the Phytozome database. The sequences were confirmed using the Smart software (Letunic and Bork, 2018), and after removing the duplicates, 71 members of the ALDH family were identified. The evolutionary relationships of the AhALDH members were analyzed using the MEGA X software (Kumar et al., 2018), while ALDH members in Arachis duranensis and Arachis ipaensis were used to correct evolutionary relationships. All these ALDH members were classified into 10 subfamilies, labelled I–X according to the order of AhALDH evolution. Subfamilies III and VI had the fewest AhALDH members, only two, while subfamily X had the highest number of AhALDH members at 17. These AhALDH members were named according to the evolutionary results and Brockers’ method (Brocker et al., 2013). ALDH members in A. duranensis and A. ipaensis, along with AhALDH members, were also divided into 10 subfamilies, which proved the accuracy of the evolutionary analysis ( Figure 1 ). Moreover, detailed information about the proteins (including protein length, molecular weight, and isoelectric point) is shown in Table S1 . The protein length varied from 303 to 791 amino acids, the molecular weight was 32.48–88.83 kDa, and theisoelectric point was 4.91–9.61. This information revealed that AhALDH members in different subgroups had diversified features, indicating that different members might perform different functions.

Location information was obtained from databases and drawn using Tbtools ( Figure 2 ). All chromosomes had AhALDH members, except for Chr 10 and Chr 20. The number of AhALDH members on each Chr differed: Chr13 and Chr15 had the most AhALDH members (9), while Chr03 and Chr11 had seven members; Chr1, Chr7, and Chr12, with only one member each, had the fewest, while only two AhALDH members were found on Chr16 and Chr19. These results indicated that the distribution of AhALDH members on groundnut chromosomes is not uniform, indicating that Chr13 and Chr15 might be central to AhALDH evolution.

The motif and gene structure of AhALDHs were analyzed using MEME and GSDS, respectively, to explore their structure and properties. The results are shown in Figure 3 . The length of the motifs varied from 10 to 50, and the e-value was lower than 1e⁻²⁰. AhALDHs in the same subfamily shared similar kinds of motifs, although some motifs also varied within the same subfamily, such as motif 1 ( Figure 3 ). Members of the same subfamily had a similar gene structure, with the longest AhALDH member being in subfamily X. Moreover, the motif and gene structure of ALDH members in A. duranensis and A. ipaensis also revealed that members in each subfamily had similar characteristics, indicating that they might perform similar functions.

In this study, 260 ALDH members in seven species, namely groundnut, Arabidopsis, rice, human, cotton, soybean, and grape (Vitis vinifera), were used for the evolutionary and motif analysis. These 260 members were divided into 10 subfamilies, in which the members within each subfamily shared similar motifs. The evolutionary results were similar to those shown in Figure 2 . Under the same analysis threshold, the AhALDHs had similar kinds of motifs to ALDH members in other species ( Figure 4 ).

The cis-acting elements of AhALDHs were analyzed and their function predicted using the Plantcare software ( Figure 5 ; Table S3 ). Three kinds of elements were recognized based on their function. The cis-acting elements TGA-element, ABRE, P-box, GARE-motif, TCA-element, AT-rich sequence, SARE, TATC-box, and AuxRR-core (marked in red in Figure 4 ) were predicted to be related to hormones; LTR, MBS, ARE, and GC-motif (marked in blue) were related to abiotic stress. The members within each subfamily had similar cis-acting elements, and the analysis revealed that AhALDHs might be associated with hormones and abiotic stress; in particular, some members had stress-related cis-acting elements.

Fifty-two pairs of collinear AhALDHs were identified. Most of the AhALDH members had more than two collinear pairs; AhALDH2B3 had five collinear pairs, which revealed that AhALDH2B3 might have a longer and more pivotal evolutionary relationship ( Figure 6A ). Compared with the Arabidopsis members, AhALDHs had 24 collinear pairs and likely similar functions to those of Arabidopsis collinear genes. Interestingly, both collinear Arabidopsis members and AhALDH members were not distributed on the last chromosome, which potentially reveals some characteristics of the ALDH members ( Figures 2 , 6B ).

The GO and KEGG enrichment analyses ( Figure 7 ) were conducted to understand the function of AhALDHs. The AhALDHs were significantly (P < 0.05) enriched in 81 GO terms ( Table S4 ); the top-10 GO terms were enriched in semialdehyde dehydrogenase activity (6) and amino acid-related terms (4), suggesting that AhALDHs might participate in the growth of plants ( Figure 7A ). The analysis further revealed that the AhALDHs were significantly (P < 0.05) enriched in 20 pathways ( Table S5 ); eight pathways were related to amino acid metabolism, indicating the role of AhALDHs in the biosynthesis of amino acids ( Figure 7B ). These results indicated that AhALDHs might exercise function through amino acid pathways.

The expression pattern of 28 AhALDH members from different tissues encompassing all subfamilies was extracted from the Phytozome database, and a heatmap was drawn using TBtools. The results showed that the expression of AhALDHs was tissue-specific ( Figure 8 ). Some AhALDHs had a higher expression level (such as AhALDH12A1, AhALDH18B1, AhALDH6B2, AhALDH2C7, AhALDH6B4, AhALDH2C11, and AhALDH11A5) in roots than in other tissues; therefore, roots should be used as a target tissue of AhALDHs for further study.

The ALDH activity was determined under control and stress conditions at 0, 12, 24, 48, 72, and 96 h ( Figure 9 ). The ALDH activity showed no obvious trend under the control conditions, but the data of activity under saline-alkali stress had increased. The results showed that the ALDH activity began to significantly change at 48 h (P < 0.05). All these results illustrated that 48 h could be used as an simulated stress time for testing the expression of AhALDH members according to ALDH Activity.

Twelve AhALDHs with higher expression in the roots at the sprout stage, based on the tissue-specific expression pattern analysis, were selected for qRT-PCR analysis under stress, and 48 h was used as the sampling time. The expression of these members significantly differed between the control and stress treatment (P < 0.05). Some AhALDH members, AhALDH10A1, AhALDH22A1, AhALDH12A1, AhALDH6B2, AhALDH3H2, AhALDH3H1, AhALDH10A4, AhALDH2C11, AhALDH11A5, and AhALDH3H3, were upregulated under saline-alkali stress, while others, AhALDH22A4 and AhALDH3H4, were downregulated. These results suggested the role of these members in plant response to abiotic stress; some might be involved in positive regulation, while some might be involved in negative regulation ( Figures 10A-L ).