*3.10. Protein–Protein Interaction Network Analysis*

The protein network interaction analysis can help understand protein biological function's and mechanisms [71]. Since both the RNA-seq data and GO annotation analysis suggested the role of HSFs in stress conditions and normal growth, we performed network analysis to predict the interacting partners of *Zm*HSFs (Table S7). The results showed that maize HSFs interact with themselves and a range of proteins with well-known functions in cellular growth and stress responses (Figure 12). For example, HSFs were shown to interact with molecular chaperons HSP101, HSP82 (belongs to HSP90 family), HSBP-2, and DnaJ-like protein (belongs to the HSP40 family). It was reported that HSP101 and HSA32 interact with each other and promote acquired thermotolerance in *Arabidopsis* [72]. The HSP82 was reported to be induced by higher temperatures. A higher concentration of HSP82 is required for normal cellular growth in yeast at higher temperatures [73]. Gu et al. reported that maize HSBP-2 and HSFA2 interact with each other and modulate raffinose biosynthesis [74]. HSFA2 was shown to bind to the promoter sequence of HSBP-2 and

activate its expression. Higher raffinose synthesis improved HS tolerance of *Arabidopsis thaliana*. The DnaJ-like proteins are molecular co-chaperones that interact with HSP70s and control protein homeostasis [75]. DnaJ proteins have been reported to play a critical role in plant growth, development, and HS tolerance [75–77]. *Zm*HSFs also interact with two major proteins, i.e., multi-protein bridging factor 1c (MBF1c) and DREB2A. Both these proteins have been shown to accumulate under diverse abiotic stress conditions. DREB2A is a major protein, and its overexpression improves plant HS, drought stress, cold stress, etc., tolerance [78]. MBF1c is a transcriptional co-activator that modulates the expression of DREB2A, some HSFs, and phytohormones [3]. Interestingly, MBF1c is necessary for basal thermotolerance but not for acquired thermotolerance [79]. In addition, MBF1c is also shown to be required for plant developmental responses [80].

**Figure 12.** Protein–protein network of maize HSFs. The line connecting two proteins represents that an interaction exists between them.

Maize HSFs are also predicted to interact with SUMO proteins. SUMOylation is a post-translational phenomenon where SUMO proteins are covalently attached and detached to target proteins [81]. This process affects several biological processes inside the cell, including transcriptional regulation of gene expression, apoptosis, programmed cell death, cellular response to stress, stability of proteins, etc. [81]. Rytz et al. reported that SIZI, a SUMO protein, targets multiple TFs, chromatin remodelers, transcriptional coactivators/repressors connected to abiotic and biotic stress responses [82]. This suggests maize HSFs may also be SUMOylated under diverse biological conditions and stress responses. To conclude, PPI analysis aligned with the RNA-seq and GO annotation analysis, which indicated that HSFs of *Zea mays* play an important role in abiotic stress conditions and in maize growth and metabolism.

#### **4. Discussion**

Maize (*Zea mays*) is a major cereal crop that is widely cultivated worldwide for food, feed, fiber, and fuel. Maize is also considered a model plant for basic and applied research in plant science [83]. Unraveling the factors regulating the growth and stress resistance would contribute significantly to the development of climate-smart, stress-resilient maize cultivars with higher agricultural productivity. The sequencing of the maize genome (B73 inbred line) in 2009 opened a plethora of opportunities to identify, analyze, and characterize stressassociated genes in maize [84]. To provide food security in the scenario of climate change and ever-growing world population, it is imperative to understand the molecular mechanisms behind plant stress resistance and explore genetic resources associated with the higher crop yield [3,15]. HSFs have been identified in several plant species, including important crops. The *Arabidopsis thaliana*, *Oryza sativa*, *Zea mays*, *Glycine max*, *Populus trichocarpa*, *Solanum lycopersicum*, *Brachypodium distachyon*, *Sorghum bicolor*, and *Triticum aestivum* contain 21, 25, 25, 38, 28, 26, 24, 23, and 61 HSFs in their genomes, respectively [20,40–45,71,85]. Following the sequencing of several plants, it is found out that the number of HSFs may be independent of the genome size [71]. For example, *Arabidopsis thaliana* (135 Mb) contains 21 HSFs [20,86], while *Medicago truncatula* (375 Mb) harbors 15 HSFs [43,87]. Similarly, 25 HSFs are found in *Oryza sativa* (430 Mb) [40,88] and an equal number of HSFs are also present in *Zea mays* (2.4 Gb) [41,84]. Even though the HSF gene family was previously characterized by Lin et al. [41], our work differs from theirs in multiple aspects. Their research was mostly restricted to the identification and classification of *Zm*HSFs. On the other hand, this comprehensive study particularly focused on the evolutionary analysis, expression profiling, GO, and PPI networks to explore the probable regulatory role played by *Zm*HSFs under benign and stress conditions.

The distribution of the HSF gene family in maize was analyzed by constructing a chromosomal map (Figure 1). The fact that all the chromosomes harbor at least one HSF gene suggests that *Zea mays'* most recent common ancestor has HSF genes distributed widely in its genome. Phylogenetic analysis indicated the *At*HSFA2, *At*HSFC1 did not align with sub-class A2, C1 in the present study, which aligns with the results reported by Lin et al. [41]. Maize HSFs are divided into three classes and further into 13 sub-classes which is consistent with the HSF class number observed in other monocots. For example, the HSFs of *Oryza sativa*, *Brachypodium distachyon*, *Sorghum bicolor*, and *Triticum aestivum* are also divided into three classes and 13 sub-classes [40,45,46,85]. Despite that, differences among HSF numbers were observed in different sub-classes between monocots (Figure 2, Table 2). For example, compared to rice, *Brachypodium*, and *Sorghum*, the sub-families B4, A2, and C2 contain fewer HSF members in maize. On the other hand, the sub-classes A1, A4, A8, B1, and B2 are expanded in *Zea mays* (Figure 2, Table 2). Gene duplications generate new genes and provide novel possibilities for evolutionary success [89,90]. In fact, it has been proposed that tandem and segmental duplications have been the primary driving source of evolution as these events lead to expansion of gene families and generation of proteins with novel functions [91]. Tandem duplication involves the duplication of two or more genes located on the same chromosome, while segmental duplication refers to the phenomenon when genes belonging to the same clade but located on different chromosomes are duplicated [92]. In the present analysis, nine pairs of *Zm*HSFs were shown to be paralogs (Table 3, Figure 3). The results indicate that segmental duplication events have played a major role in the expansion of the HSF gene family in maize. An increase in gene regulatory repertoire such as transcriptional regulators, developmental regulators, signal transducers, etc., is a prerequisite for the evolution of complex systems in different

organisms [22,93]. Since the gene duplication events result in the doubling of a single gene which cannot account for such large expansions, it has been suggested that whole-genome duplication (WGD) events have been instrumental in expanding the regulatory repertoire of plants [90]. It is assumed that the *Arabidopsis* genome experienced two rounds of WGD in the past 60–70 million years [94,95]. More than 90% increase in regulatory genes has been caused by duplication events *Arabidopsis* in the past 150 million years [94]. This suggests that an increase in the HSF gene members in plants accounts for WGDs. Additionally, segmental duplications occur in gene families, which evolve at a slower rate [91]. It is thought that the increase or decrease in exon number plays an important role in the evolution of a gene family [96]. Therefore, we investigated the number and distribution of introns and exons in *ZmHSF* genes. Our results showed that the *Zm*HSF genes contain 2 exons and 1 intron except for *ZmHSF-02* and *ZmHSF-24* (Figure 4; Table 1). Moreover, the length and position of exons and introns were well conserved in the same sub-classes but varied considerably between different sub-classes.

The previous investigations showed that HSFs play an important role in plant growth [39]. Therefore, we investigated the tissue-specific expression of *Zm*HSFs in 20 different developmental tissues using RNA-seq data (Figure 9A–C). Several genes showed an enhanced expression that reflects their role under various developmental stages. In particular, *ZmHSF-05* (A-2a) and *ZmHSF-06* (A-1a) were highly expressed almost across all the growth phases. The *hsfa1abde* quadruple mutants displayed abnormal phenotype and growth retardation, implying HSFA1s is also involved in developmental processes [29]. Interestingly, *HSFA2* could rescue the developmental defects of *hsfa1abde* quadruple mutants [97]. This further supports the result obtained from our analysis and provides a strong base for further wet-lab studies to characterize the function of *ZmHSF-05* and *ZmHSF-06* in plant growth and development. Similarly, HSFs have been reported to play a key role in plant acclimation to abiotic stress conditions. Kumar et al. reported that *TaHSFA6e* modulates tolerance of wheat to HS and drought stress during pollination and grain filling stages [98]. Yokotani et al. reported that *OsHSFA2e* improves Arabidopsis tolerance to HS and salinity stress by activating the expression of HSPs [99]. Thus, the expression patterns of *Zm*HSFs were evaluated under abiotic stress conditions. Most of *Zm*HSFs displayed stress-specific expression, with some HSFs showing upregulation only under particular stress events (Figure 10). Jiang et al. reported that *Zm*HSF-04 improves plant tolerance to HS, salinity stress and increases the sensitivity to abscisic acid [100]. Similarly, *ZmHSF-12* overexpression improves plant basal thermotolerance and AT [101]. These results are in line with our analysis which showed a higher transcript accumulation of these TFs under respective stress conditions (Figure 10).

The PPI analysis indicated that maize HSFs interact with molecular chaperones and stress-associated proteins (Figure 12). Molecular chaperones are present inside the cells and are constitutively expressed under normal conditions or are induced under specific developmental stages or stress conditions [102]. These chaperons perform various functions under physiological conditions inside the cells, such as signaling, folding, and stabilization, translocation, and degradation of proteins [11,102]. Under harsh environmental conditions, molecular chaperones act as powerful buffers to limit protein misfolding/unfolding and prevent protein aggregate formation that might be otherwise toxic to plant cells [25]. Here, the *Zm*HSFs were shown to interact with chaperons belonging to different families (*HSP101*, *HSP90*, *HSP40*) and genes with a well-known role in thermotolerance (*HSA32*, *HSP82*, *HSBP-2*). DREB2A is a major transcriptional activator that functions downstream of HSFA1s dependent transcriptional cascade in *Arabidopsis thaliana* [3,12,15,29]. Similarly, MBF1c is a major protein characterized by its role in regulating abiotic stress responses and growth in plants [3,82,83]. SUMO proteins are attached to their target proteins and modify their biological activities under various physiological and stress conditions [84]. In *Arabidopsis*, SUMOylation has been proposed as one of the molecular mechanisms that are responsible for the activation of HSFA1s [12]. Many HSFs in *Arabidopsis* such as *HSFA1d, HSFA2*, and *HSFB2B* have the potential to be SUMOylated [103]. In tomatoes, the knockout

of SIZI (a SUMO ligase) reduces plant thermotolerance [104]. Here, *Zm*HSFs were shown to interact with all these proteins (DREB2A, MBF1c, and SUMO proteins), which further confirms their role in regulating the abiotic stress response.

Taken together, our present analysis provides strong support for the positive role of HSFs in the growth and development of maize by the regulation of primary metabolism. Furthermore, HSFs of maize interact with the major stress-responsive proteins and confer abiotic stress resistance.
