MdHSP70 gene family identification and sequence analysisA total of 56 apple HSP70 family genes were identified using Arabidopsis amino acid sequence alignment, and named MdHSP70-1 to MdHSP70-56 based on their chromosomal locations (Fig. 1). They were located on 14 chromosomes, and the maximum number of genes located on Chr1 was 22. At the same time, obvious gene clusters appeared on Chr1, Chr7, Chr9 and Chr12. Analysis of their physicochemical properties showed that the amino acid length of the MdHSP70 gene family ranged from 58 to 891 aa, with the longest being MdHSP70-50 and the shortest being MdHSP70-32. The pI values ranged from 4.49 to 10.19 and most of them were less than 7. It is presumed that this family is mainly acidic proteins. The relative molecular weights were MdHSP70-32 (6691.45) to MdHSP70-50 (99,483.34), with α-helices ranging from 9.26 to 100%, and β-turns ranging from 0 to 12.24% (Table 1).Fig. 1Distribution of MdHSP70 genes on different chromosomes. The scale on the left indicates the chromosome length (Mb), and each chromosome is labeled with the MdHSP70 gene on the right side. Different colors indicate gene density, where red indicates high-density regions and blue indicates low-density regions.Table 1 Information and protein secondary structure analysis of MdHSP70 gene family.Phylogenetic tree analysis of the MdHSP70 gene familyThe results showed (Fig. 2) that all HSP70 family members could be categorized into seven subfamilies. According to the number of genes contained in each subfamily, The highest number of family V genes was 97, accounting for 35.80% of the total number of genes, and the lowest number of family I genes was 2, accounting for 0.74% of the total number of genes. Family I only contained the tomato HSP70 gene, and family VI only contained the MdHSP70 gene. Family III did not contain the Arabidopsis and rice HSP70 genes, and family VII did not contain the grape, soybean, and tomato HSP70 genes. Family II, VI, and V contained the nine species studied. Further analysis of their evolutionary relationships revealed that MdHSP70 family members are more closely related to Arabidopsis HSP70 family members, tomato HSP70 family members, and soybean HSP70 family members; the above results suggest that plant HSP70s are evolutionarily distinct, and that MdHSP70 family members are more closely related to dicotyledonous plants than to monocotyledonous plants.Fig. 2Phylogenetic analysis of the MdHSP70 gene family. Phylogenetic tree was constructed using HSP70 protein sequences. The NJ method was used and the bootstrap value was set equal to 1000.Covariance analysis of the MdHSP70 gene familyAmong HSP70 family members in monocotyledon plants, sorghum (31), and maize (25) contained the most and least members, respectively. Among dicotyledon plants, apple (56) and Arabidopsis (16) contained the most and least members, respectively. Therefore, the study of the HSP70 family in the above plants is helpful to understand the reasons for the expansion of gene numbers.In the monocotyledon HSP70 family, millet (Fig. 3A) has 3 pairs of collinear genes, which are distributed on Chr3, Chr5 and Chr9. Since the gene Sela.5G376100.1 is collinear with the other three genes at the same time, it is speculated that the cross replication of chromosomes in the evolution process leads to the repetition of fragments. Maize (Fig. 3B) has 4 pairs of collinear genes, which are distributed on Chr2, Chr4, Chr5 and Chr6. In rice (Fig. 3C), 4 pairs of collinear genes were distributed on Chr1, Chr3, Chr5, Chr11 and Chr12. LOC_Os05g38530 and other two genes also showed fragment repetition. Sorghum (Fig. 3D) has one pair of collinear genes, distributed on Chr1 and Chr3. In dicotyledonous HSP70 family, apple (Fig. 3E) has 4 pairs of collinear genes, which are distributed on Chr1, Chr2, Chr4, Chr7, Chr9, Chr12, Chr15 and Chr17. There are 7 pairs of collinear genes in soybean (Fig. 3F), which are distributed on Chr2, Chr5, Chr7, Chr8, Chr12, Chr13, Chr14, Chr15, Chr16, Chr17 and Chr19. Tomato (Fig. 3G) had 5 pairs of collinear genes, distributed on Chr3, Chr4, Chr6, Chr8, Chr9 and Chr10, and the gene Solyc06T00041 and Solyco3T001718 showed fragment repetition. Arabidopsis (Fig. 3H) has 3 pairs of collinear genes, which are distributed on Chr1, Chr3, Chr4 and Chr5. In grape (Fig. 3I), 5 pairs of collinear genes were distributed on ChrUn, Chr3, Chr6, Chr8, Chr9 and Chr13, and fragment repetition occurred in genes VIT_213s0019g01430, VIT_208s0007g00130 and VIT_206s0004g04510. It is noteworthy that all nine species studied showed clusters of genes at chromosomal locations. The above results suggest that the HSP70 gene family has a high degree of homology, in which segmental duplications may play an important role in the expansion of the family.Fig. 3Intraspecic collinearity analysis of HSP70 gene family. (A)–(I) Represent the collinear heat maps of species millet, maize, rice, sorghum, apple, soybean, Arabidopsis, tomato and grape. The blue line shows monocotyledonous plants and the red line shows dicotyledonous plants.The collinearity analysis among monocotyledonous leaves of rice, millet, sorghum and maize showed (Fig. 4A) that there were 49 gene pairs between rice and millet, 30 gene pairs between rice and sorghum, 16 collinear pairs between rice and maize, 31 gene pairs between millet and sorghum, and 17 collinear pairs between millet and maize. There were 15 collinear pairs between sorghum and maize, indicating that the relationship between rice and millet was closer, followed by millet and sorghum, and the least collinear pairs between sorghum and maize, indicating that the relationship between sorghum and rice was estranged; Collinear analysis of dicotyledon apple, Arabidopsis, tomato, grape and soybean HSP70 family members found (Fig. 4B) that there were 36 gene pairs between apple and Arabidopsis, 28 gene pairs between apple and tomato, 17 gene pairs between apple and grape, 16 gene pairs between apple and soybean, and 31 gene pairs between Arabidopsis and tomato. There are 16 gene pairs between Arabidopsis and grape, 14 gene pairs between Arabidopsis and soybean, 15 gene pairs between tomato and grape, 15 gene pairs between tomato and soybean, and 8 gene pairs between grape and soybean. The results indicated that apple and Arabidopsis were more closely related, while grape and soybean were more closely related.Fig. 4Interspecific collinearity analysis of HSP70 gene family. (A) The collinearity of the HSP70 gene in monocotyledonous plants such as rice, millet, sorghum and maize. (B) The collinearity of HSP70 gene in dicotyledonous plants such as apple, Arabidopsis, tomato, grape and soybean. The blue lines represent collinear genes.MdHSP70 family protein interaction and screening and validation of proteins interacting with MdDVH24_032563To determine the interactions between MdHSP70 and related proteins, we constructed an HSP interaction network using a STRING database with apple proteins. The results showed that there were interactions among the 20 MdHSP70 proteins, and the interactions among the 20 MdHSP70 proteins formed a protein interaction network (PPI network) (Supplementary Fig. S1). It is worth noting that DVH24_004410 and DVH24_032563 are proteins containing the OUT domain, which belong to the UDP-glycosyltransferase family (UGt). Since the UDP-glycosyltransferase family plays an important role in the regulation of plant growth and abiotic stress tolerance, and the thickness of the lines represents the intensity of the interaction. We can find that MdHSP70-43, MdHSP70-14, MdHSP70-35, MdHSP70-11, MdHSP70-10, MdHSP70-53, MdHSP70-50, MdHSP70-8, and MdHSP70-13 are the major interacting proteins compatible with MdDVH24_004410, MdDVH24_032563, and it is speculated that these proteins play an indispensable role in improving the growth and development of apple plants and abiotic stress tolerance. In order to verify the interactions between the proteins even further, we constructed MdDVH24_032563-pGBADT7, MdHSP70-53-pGBKT7 vectors, cloned the above genes and transformed them into E. coli DH5α (Supplementary Fig. S2). Co-transformation of pGBADT7 with pGBKT7, pGBADT7 with MdHSP70-53-pGBKT7, MdDVH24_032563-pGBADT7 with pGBKT7, and MdDVH24_032563-pGBADT7 with MdHSP70-53-pGBKT7 into Y2H yeast-receptor cells was performed to Validation. AD/BD, AD/MdHSP70-53-BD, and MdDVH24_032563-AD/BD grew normally in DDO and DDO/X media and did not blotch or turn blue on QDO and QDO/X, indicating that MdHSP70-53-BD and MdDVH24_032563-AD are not self-activating. MdDVH24_032563-AD/MdHSP70-53-BD appeared as white spots on QDO and QDO/X as well as DDO and DDO/X media and turned blue, suggesting the presence of an interaction between MdDVH24_032563 and MdHSP70-53 (Fig. 5).Fig. 5Y2H analysis of MdHSP70-53 protein. AD and BD represent pGADT7 and pGBKT7, respectively. DDO indicates SD/-Leu/-Trp, and QDO indicates SD/-Ade/-His/-Leu/-Trp. X indicates X-α-gal.Analysis of cis-acting elements of the MdHSP70 gene familyCis-acting elements are non-coding DNA sequences that regulate transcriptional initiation of related genes. In this study, PlantCARE software was used to analyze the 2000 bp promoter sequence upstream of the MdHSP70 coding region, and a total of 9 cis-acting elements in response to hormone signals and stress responses were identified (Fig. 6). Hormone response elements include MeJA, SA, indole-3-acetic acid (IAA), gibberellin (GA) and ABA, and light responsive cis-elements (LREs). The cis-regulatory factors related to abiotic stress include low-temperature response elements, drought-induced response elements and stress-related elements. The promoter region of MdHSP70-15 contains only GA response elements, and the promoter region of MdHSP70-56 contains only SA response elements. Most of the genes contain MeJA response elements, GA response elements, ABA response elements and drought induced response elements.Fig. 6Cis-acting element analysis of the first 2000 bp of the MdHSP70 gene family. Different elements are marked with different colors. Numbers indicate the number of cis-acting elements.Analysis of MdHSP70 gene family structure and conserved motifs and conserved domainsThe gene structure of 56 MdHSP70 gene family genes was arranged according to the phylogenetic tree, and the analysis results showed (Fig. 7) that the MdHSP70 gene family was divided into three groups according to the evolutionary tree relationship. The first group is composed of MdHSP70-1, MdHSP70-32, MdHSP70-23, MdHSP70-46, MdHSP70-52, MdHSP70-53, and MdHSP70-33, except that MdHSP70-32 and MdHSP70-52 contain only motif1. Other genes contained motif1, motif2, motif3, motif4 and motif6. The second group consists of 45 genes, almost all containing motif5 and motif9 except for a few genes. The third group of genes consists of MdHSP70-29, MdHSP70-36, MdHSP70-31 and MdHSP70-49, and it contains motif3. The N-terminal starts with motif1 and motif7, and the C-terminal ends with motif8. It is worth noting that motif1, motif3 and motif7 appeared frequently in the entire MdHSP70 gene family, suggesting that motif1, motif3 and motif7 are important and highly conserved domains of this gene family. The number of introns and exons of MdHSP70 gene in different subfamilies is different, and the exon number is between 1 and 16. MdHSP70-24 has the highest number of exons, with 16. The exon length of MdHSP70-22 is the longest, about 16500 bp, and the exon length of MdHSP70-32 is the shortest. The number and distribution of introns and exons within the same group also are different, with only a few having highly conserved properties. The conserved structural domains of the MdHSP70 gene family show that except for the MdHSP70-24 gene, the conserved structural domains of all genes are HSP70.Fig. 7MdHSP70 gene family structure and conserved motif analysis. (A) Exon–intron structure of MdHSP70 gene. (B) Conserved motifs; different motifs are marked with different colors and the numbers above represent different motifs. (C) Conserved structural domain of the MdHSP70 genes.Tissue-specific expression analysis of MdHSP70 gene familyThe expression of 47 MdHSP70 genes in different plant organs at different periods was analyzed by searching the Apple MDO database, expressed value is FPKM (Supplementary Fig. S3). Comparing the expression of these genes in apple seed, stem, bud, leaf, flower and fruit stages, these genes can be categorized into three classes. Class I contained only MdHSP70-11 gene, which were highly expressed in almost all four periods except for low expression at flowering stage, and the highest expression was found at leaf and fruit stages, indicating that the genes in this class might be related to leaf growth and fruiting in apple, and their role in fruiting was more obvious. Class II contained 33 genes which were expressed at lower levels in five periods. Class III contained 9 (MdHSP70-23, MdHSP70-27, MdHSP70-37, MdHSP70-44, MdHSP70-45, MdHSP70-47, MdHSP70-49, MdHSP70-50, MdHSP70-53) genes, and these genes were more highly expressed in some periods. We found that 16 MdHSP70 genes were expressed at the seed stage, accounting for about 34% of the total retrieved genes, and the expressions of MDHSP70-1 and MDHSP70-47 were the highest; twenty genes were expressed at rooting stage, accounting for 43% of the total retrieved genes, and the expression of MdHSP70-1 was the highest. In the germination stage, the expression of MdHSP70 increased gradually with the increase of time, MdHSP70-1, MdHSP70-23, MdHSP70-37, MdHSP70-47, MdHSP70-50 were the most obvious. In the leaf growth period stage, MdHSP70 increased first and then decreased with the increase of time. The expression of MdHSP70 in flowering stage was lower than that in germination stage. The expression level of MdHSP70 in flowering stage was lower than that in leaf growth period stage: the expression level of MdHSP70 in fruit stage showed a trend from increasing to decreasing, then increasing and decreasing again. As a whole, MdHSP70-1, MdHSP70-23, MdHSP70-27, MdHSP70-37, MdHSP70-47 and MdHSP70-50 participated in all stages of apple growth and were highly expressed, suggesting that these genes played an important role in apple growth and development.Codon preference analysis of MdHSP70 gene familyThe frequency of relative synonymous codons in MdHSP70 genome was analyzed (Supplementary Table S1). It was found (Fig. 8A) that RSCU ≥ 1 for 30 codons: AAG, UCU, CCU, CUU, GGU, GUG, CCA, GUU, AUU, ACU, UGA, GCU, UUG, AGG, GCC, AAC, AUG, UUU, CAU, UCA, UAU, UGU, GAG, GAU, ACC, CAG, AGU, AGC, AUC, AGA, among the 30 codons, 14 of them have the third codon U, accounting for 26.4% of the total codons. Seven of the third codons are G, accounting for 15.2% of the total codons, five of the third codons are C, accounting for 7.1% of the total codons, and four of the third codons are A, accounting for 5.3% of the total codons, indicating that the MdHSP70 protein genome codons prefer ending in U or G (Fig. 8B). The average values of CAI, CBI, Fop, and NC of MdHSP70 family members are 0.209, − 0.030, 0.391, and 48.750, respectively. The content of GC of MdHSP70 family members ranges from 39 to 56%, and the content of GC3s ranges from 30 to 78%. The average for GC and GC3s was 42.7% and 44.6% respectively (Supplementary Table S2). A total of 18 genes had NC values less than 50, which were: MdHSP70-1, MdHSP70-8, MdHSP70-12, MdHSP70-16, MdHSP70-18, MdHSP70-19, MdHSP70-26, MdHSP70-28, MdHSP70-29, MdHSP70-32, MdHSP70-36, MdHSP70-38, MdHSP70-42, MdHSP70-46, MdHSP70-49, MdHSP70-52, MdHSP70-53, MdHSP70-54. This indicates a strong codon preference for these 18 genes (Fig. 8C). The correlation analysis results of MdHSP70 genome codon using bias correlation indicators (Fig. 8D) showed that there is a very significant (P < 0.001) positive correlation between GC value and GC3s value, indicating that there is little difference between the first and third base of MdHSP70 genome codon. The GC values were not significantly (P < 0.05) correlated with the CBI and FOP values, and the GC3s values were not significantly (P < 0.05) correlated with the CBI and FOP values, suggesting that codon base composition has a small effect on codon usage bias.Fig. 8Analysis of the use and number of synonymous codons associated with codons in the MdHSP70 gene family. (A) Use of synonymous codons. (B) Preferred number of synonymous codons. (C) Analysis of MdHSP70 gene family codon parameters. “A3s, G3s, C3s, T3s” refers to the frequency of the third base corresponding to the synonymous codon; “CAI” is the codon adaptation index; “CBI” refers to the codon bias index; “FOP” is the frequency of optimal codon occurrence; “ENc” is the number of effective codons; “GC3s” is the number of the third codon (G + C); and GC is the gene count (G + C). (D) Correlation analysis of MdHSP70 gene family codons. Blue color indicates a positive correlation, red color indicates a negative correlation, and white color indicates no correlation. The darker the color and the larger the circle, the stronger the correlation, and vice versa. The number of observations (n) for the correlation coefficient is 141.qRT-PCR analysis of MdHSP70 gene familyThe results of quantitative fluorescence analysis of apple leaves treated with different conditions showed that (Fig. 9). 41 MdHSP70 gene expressions were up-regulated under both hormone and abiotic stress treatments. Under hormone treatment conditions, MdHSP70-23, MdHSP70-26, and MdHSP70-45 had the highest expression under MeJA, GA3, and ABA treatments, respectively, which were 44.36-fold, 43.96-fold, and 51.73-fold higher than the control. Under abiotic stress treatment conditions, MdHSP70-45 had the highest expression after NaCl treatment, which was about 91.70-fold of the control; MdHSP70-6 had the highest expression after PEG treatment, which was about 95.30-fold of the control; and MdHSP70-45 had the highest expression after 4 °C treatment, which was about 93.82-fold of the control.Fig. 9Expression levels of MdHSP70 gene family under abiotic stress and plant hormone treatment. Statistical analysis is performed by one-way ANOVA with Tukey’s honestly significant difference (HSD) test. The control unstressed expression level, which was assigned with a value of 1. Error bars represent the mean ± SE from three biological repeats. Different letters denote signifcant differences, whereas the same lowercase letters indicate no statistical difference (P < 0.05).The expression of the remaining members was down-regulated under individual treatments, for example, MdHSP70-10, MdHSP70-14, MdHSP70-16, MdHSP70-18 and MdHSP70-56 gene expressions were down-regulated after MeJA treatment; MdHSP70-14, MdHSP70-34, and MdHSP70-39 were down-regulated in gene expressions after GA3 treatment; and MdHSP70-14, MdHSP70-16, MdHSP70-34, and MdHSP70-39 were down-regulated in gene expression after ABA treatment. Likewise, under abiotic stress treatments, the expression of MdHSP70-3 and MdHSP70-10 were down-regulated after PEG treatment, and MdHSP70-2 and MdHSP70-10 gene expressions were down-regulated after low temperature treatment at 4 °C. It is noteworthy that most MdHSP70 genes were more sensitive to stress treatments relative to hormone treatments, suggesting that MdHSP70 genes play an important role in abiotic stresses in plants. The relationship between MdHSP70 and hormones and abiotic stress remains to be verified in our further experiments.Prediction of transmembrane structural domains and subcellular localization analysis of the MdHSP70 geneSeven genes were found to have transmembrane structures on the Transmembrane Structure Domain Prediction website. They are MdHSP70-4, MdHSP70-13, MdHSP70-28, MdHSP70-30, MdHSP70-35, MdHSP70-43, and MdHSP70-50 (Supplementary Fig. S4). Whereas, MdHSP70-6, MdHSP70-26, and MdHSP70-45 do not have transmembrane structures, suggesting that they are localized only to a organelle. Subcellular localization can be used to detect the exact location in the cell where specific proteins are present and act. They are transiently expressed using recombinant plasmids ligated with fluorescently labeled vectors, and their location is determined by observing fluorescence in the cell through an inverted microscope. Subcellular localization is not only used to study the structure and function of cells, but can also be used to study cell signaling, the interactions that exist between proteins, and to track the location and activity of cell surface receptors. To further determine the subcellular location of MdHSP70, we cloned the genes of MdHSP70 (6, 26 and 45) and constructed the overexpression vectors 35S::MdHSP70-6-GFP, 35S::MdHSP70-26-GFP, 35S::MdHSP70-45-GFP, and the empty vector 35S::GFP was used as a positive control to observe green fluorescence. The identification results showed that the overexpression vectors 35S::MdHSP70-6-GFP, 35S::MdHSP70-26-GFP, and 35S::MdHSP70-45-GFP observed green fluorescence in the cell membrane (Fig. 10).Fig. 10Subcellular localizations of MdHSP70-6, MdHSP70-26, and MdHSP70-45 in tobacco leaves. 35S::GFP was used as the control.
