Skip to main content
Journal of Genetic Engineering and Biotechnology
Download PDF

Protein-protein interaction (PPI) network analysis reveals important hub proteins and sub-network modules for root development in rice (Oryza sativa)

Journal of Genetic Engineering and Biotechnology volume 21, Article number: 69 (2023) Cite this article

Abstract

Background

The root system is vital to plant growth and survival. Therefore, genetic improvement of the root system is beneficial for developing stress-tolerant and improved plant varieties. This requires the identification of proteins that significantly contribute to root development. Analyzing protein-protein interaction (PPI) networks is vastly beneficial in studying developmental phenotypes, such as root development, because a phenotype is an outcome of several interacting proteins. PPI networks can be analyzed to identify modules and get a global understanding of important proteins governing the phenotypes. PPI network analysis for root development in rice has not been performed before and has the potential to yield new findings to improve stress tolerance.

Results

Here, the network module for root development was extracted from the global Oryza sativa PPI network retrieved from the STRING database. Novel protein candidates were predicted, and hub proteins and sub-modules were identified from the extracted module. The validation of the predictions yielded 75 novel candidate proteins, 6 sub-modules, 20 intramodular hubs, and 2 intermodular hubs.

Conclusions

These results show how the PPI network module is organized for root development and can be used for future wet-lab studies for producing improved rice varieties.

Background

Oryza sativa has a very high demand as a staple food, and although much successful research has been carried out to improve the yield, it tends to decrease drastically in response to environmental stresses. Therefore, the improvement of O. sativa must continue, aiming at sufficient supply for the increasing population, which requires producing improved O. sativa varieties with higher yields and a higher ability to withstand environmental stresses [1,2,3].

The root system is the main component that supplies water and nutrients to the plant and plays a major role in withstanding environmental stresses [4]. Therefore, it should have a high priority when improving plant varieties. Root development is a complex biological process regulated by a collection of biological pathways, which are influenced by environmental and genetic factors [4]. This research is focused on investigating the genetic factors by identifying the functionally important proteins and their interactions responsible for root development in O. sativa using network analysis.

Biological processes are regulated by a collection of proteins and their interactions. These protein-protein interactions (PPI) are represented as PPI networks [5, 6]. PPI data are generated using wet-lab and computational techniques and are stored in databases [7, 8], such as DIP, STRING, and BioGRID. Among these databases, the STRING database is popular because of the higher abundance, coverage, and better quality control of PPI data [7, 9, 10]. STRING contains PPIs from both experimental computational methods and provides a combined quality score for each interaction by integrating the data from various resources such as literature and gene expression profiles [7, 10, 11]. Using this approach produces more accurate predictions compared to networks that solely rely on experimentally determined interactions, which are prone to high rates of false positives and false negatives in identifying interactions [10, 12]. PPI networks contain modules, which are distinct collections of proteins usually specific for a particular function or a phenotype [5, 13]. Hence, PPI networks can be analyzed to identify modules, which represent underlying protein interactions that determine the molecular functions and phenotypes. Furthermore, PPI networks are used for predicting novel protein candidates for molecular functions and phenotypes based on their interactions with known neighbors [14, 15]. Though wet-lab methods are available for predicting new protein candidates, computational methods for protein function prediction are faster, more cost-effective, and less laborious than wet-lab methods [16, 17]. Sequence similarity-based methods are popular computational approaches, which have been proven to be effective for some protein molecular function prediction studies, but they are less efficient for phenotype studies [16, 18]. This is because proteins associated with one phenotype can include proteins with highly diverse sequences, annotated with different molecular functions [12, 15, 18]. Therefore, predicting proteins for phenotypes using PPI networks is more accurate and comprehensive than sequence similarity-based methods [16, 18].

PPI network analysis can be used to identify the sub-modules within a particular module of a phenotype, and analysis of these sub-modules allows one to identify the related biological pathways and important proteins, i.e., hub proteins, involved in those pathways [19, 20]. Identifying the hub proteins of a module is an important advantage of performing network module analysis [15]. Hubs have a higher number of interactions compared to non-hubs [21]. There are two types of hubs: intramodular and intermodular hubs [22]. Intramodular hubs can be found with their partners within the functional modules, while intermodular hubs act between the modules and interconnect them [21,22,23]. Removal of a hub has a higher impact compared to non-hubs because it impacts several biological pathways in the network [22], which disrupts the resulting phenotype. Therefore, hub proteins are identified as important proteins that play a critical role in maintaining module organization and stability [24]. These are usually important drug targets in human-related studies [25] or genetic engineering targets in crop improvement [26].

PPI networks allow the understanding of the global organization of PPIs, sub-modules, connectivity among those sub-modules, and the hub proteins [8, 27, 28]. The interpretation of these networks reveals the biological pathways associated with a particular phenotype. The efficiency of PPI network analysis has been proven in human disease research [8, 29, 30], but to our knowledge, this method has never been used on root development in O. sativa.

This study aimed to apply PPI network analysis techniques used in other biological fields to analyze proteins involved in root development in O. sativa. This involved predicting potential novel candidates for root development, visualizing and identifying sub-modules, and analyzing their biological pathways. Additionally, this study also aimed to identify potentially important hub proteins related to root development and characterize the key interactions that are related to this function. The results shed light on how protein-protein interactions (PPIs) play a crucial role in root development, which can be useful for future studies aimed at improving root architecture and developing more stress-resistant varieties of O. sativa.

Methods

Data retrieval and preprocessing

The proteins already known to be involved in root development (seed proteins) were retrieved from the literature and the STRING database (version 11.0; July 2019; https://string-db.org/). The PPI network and supplementary data for O. sativa were downloaded from the STRING database (retrieved and downloaded on July 17, 2019).

To improve the reliability and the quality of the downloaded PPI network, it was filtered using the “combined score” according to a recommended cutoff mark of 400 [10]. Duplicates of the same record were removed, and STRING identifiers (IDs) for proteins were converted to preferred protein names to facilitate further analysis.

Network-based candidate gene prediction and root development protein module extraction

The Hishigaki method was selected for the candidate gene prediction [14], and the prediction score was calculated according to the equation below.

$$prediction\, score= \frac{{\left({n}_{f\left(u\right)}- {e}_{f}\right)}^{2}}{{e}_{f}}$$

Here, f denotes the function of interest, and u denotes the protein of interest. The number of proteins with the function (f) in the n-neighborhood of u is given by nf(u), and ef denotes the expected frequency for the function calculated as follows:

$$ef=\frac{{tot}_{f} \times n(u)}{{tot}_{n}}$$

The total number of proteins annotated to the function of interest (f) in the network is denoted by totf, and totn denotes the total number of proteins in the network; n(u) denotes the total number of proteins in the immediate neighborhood of the interested protein (u) [14].

Predicted scores were sorted to obtain the proteins with the highest predicted scores. The top 20, 50, 75, and 100 proteins were listed with seed proteins, and PPI modules for those lists were extracted from the preprocessed network and visualized using the Cytoscape software (version 3.7.1) [31, 32]. The final cutoff used for the rest of the analysis was chosen to maximize the number of seed proteins that were captured by the algorithm as well as to minimize the number of possible false positives that may result from less stringent cutoffs.

Validation of the predictions

Computational validation of the predicted proteins was required to confirm the accuracy of the predictions. Validation was done using enrichment analysis and performing a literature search on the predicted proteins.

Enrichment analyses were performed using the DAVID (DAVID Bioinformatics Resources 6.8; https://david.ncifcrf.gov/home.jsp) web application. The functional annotation tool in DAVID was used, and the official gene symbol was selected as the identifier [33]. The biological process component of the Gene Ontology (GO-BP) and KEGG pathways, which had significant p-values (< 0.05), was selected for further analysis [33]. Literature searches were also used to further validate the predictions and the enriched biological pathways.

Identification and analysis of sub-modules

Preliminary identification of sub-modules was done using MCODE (version 1.5.1) [34, 35] plug-in in Cytoscape software with the clustering parameters as follows: degree cutoff = 2, node score cutoff = 0.6, k-core = 2, and max. depth = 100, and further cluster expansions were done by observing the network module visualization.

Enrichment analysis and functional interpretation of sub-modules and hub proteins were performed using the DAVID enrichment analysis tool (DAVID Bioinformatics Resources 6.8) and literature mining.

Identification and analysis of hub proteins

Intramodular hub proteins were selected according to the degree of each protein. The degree cutoff for hub selection was determined by analyzing the degree distribution and picking the top 10% of proteins with the highest degrees [21]. Furthermore, intermodular hubs were identified by analyzing the inter-modular connections which connect different sub-modules. Specifically, the proteins which connect at least 3 sub-modules were selected as intermodular hubs. Functional interpretations of hub proteins were performed by investigating the literature.

The methodology of this study is briefly illustrated in Fig. 1.

Fig. 1
figure 1

PPI network-based candidate protein prediction and validation workflow

Full size image

All required scripts were written in Python (version 2.7) and deposited in a GitHub repository (https://github.com/Samadhi9/PPIN-analysis).

Results

Data retrieval and data preprocessing

Altogether, there were 51 seed proteins extracted from the literature and the STRING database (Supplementary Table S1). The O. sativa STRING PPI network contained 25,106 proteins and 8,949,048 interactions. There were 21,212 proteins and 1,608,106 interactions after filtering by > 400 combined score cutoff. The number of interactions was reduced to 803,817 after removing duplicates.

Network-based candidate protein prediction and root development network module extraction

The Hishigaki method was used for network-based candidate protein prediction, and after several trials and errors, the top 75 candidates were selected as the most suitable number of candidates for further analyses because it gave the best visualization of the root development network module by connecting most of the sub-modules. Moreover, a significant number of seed proteins were included in the extracted network module (Table 1).

Table 1 The number of seed proteins that were present and absent in the root development network modules according to different thresholds of selecting protein candidates
Full size table

As shown in Table 1, the top 20 and the top 50 candidates had a lower number of seeds present compared to the top 75 and 100. The top 75 and 100 had better seed retention, and both retained the same number (45) of seed proteins. Furthermore, degree distribution plots for the modules with 75 and 100 candidates showed an overlap with insignificant fluctuations (Fig. 2). Therefore, the top 75 proteins were selected as the cutoff, as the networks generated from the top 75 and 100 both retained the same overall structure (i.e., had the same degree distribution), and the top 75 network would contain less false-positive results due to the lower number of predictions.

Fig. 2
figure 2

Degree distribution comparison between modules containing top 75 and 100 predicted proteins

Full size image

Although selecting the top 100 candidates connects more proteins in the network module, visualization was not clear due to congestion, and it did not reveal novel information about new sub-modules (Supplementary Fig. 1). It was just an expansion of the existing sub-modules, which caused the integration of the 3rd and the 4th sub-modules shown in Fig. 3. Although the sub-modules 3 and 4 are integrated as in Supplementary Fig. 1, according to the enrichment analysis results, they are involved in different pathways: cytokinin-activated signaling pathway and cell wall organization, respectively. Therefore, top 75 protein candidates were selected for further analyses. Their prediction scores are given in Supplementary Table S2.

Fig. 3
figure 3

PPI network module for root development in O. sativa including the top 75 predicted proteins. Teal nodes represent the seed proteins, and yellow nodes represent the predicted proteins. Edges represent the interactions between proteins. The size of the node is proportional to the degree of the node. The numbered circles separate the sub-modules

Full size image

Root development network module visualization

Visualization of the root development PPI network module with the top 75 predicted proteins is given in Fig. 3.

Altogether, there were 120 proteins including 75 predicted candidates and 45 seed proteins (Supplementary Table S2) (Fig. 3). However, 6 seed proteins (Supplementary Table S3) were not visualized in the network because 3 of those were not included in the STRING raw dataset, and 3 of those were removed while data preprocessing (Supplementary Table S3).

Computational validation of the predictions

The enriched GO-BP terms resulting from the functional enrichment analysis of the predicted root development protein candidates are shown in Table 2. To validate their expression patterns, transcriptomic data was retrieved from the NCBI database from the project titled “Transcriptome profiling of various organs at different developmental stages in rice” (BioProject ID: PRJNA243371). The root samples for the transcriptional profiling done in the abovementioned dataset were taken before and after flowering [36].

Table 2 Enriched GO-BPs terms for predicted proteins that are related to root development and the pattern of expression of the corresponding genes in the root
Full size table

The GO term: auxin-activated signaling pathway (GO:0009734) is the most enriched term for the predicted protein list. Auxin is a growth coordinator hormone that regulates where, when, how much, and what sort of growth should occur in a plant. Auxins are expressed in many parts of the root such as root tip, root cap, and root epidermis. Moreover, auxins can be seen in the primary root, lateral roots, and root hairs. They are essential for many development processes in the root such as fine-tuning the growth rates, gravitropic root growth, cell division, proliferation, differentiation, and elongation of the root [37, 38].

The biological process cytokinin-activated signaling pathway (GO:0009736) was also enriched according to Table 2. Cytokinin plays several roles in root development, including regulating the responses to the growth nutrients and the biotic and abiotic stresses. Furthermore, cytokinin regulates root differentiation, elongation, branching, and root architecture [38, 39]. Moreover, it inhibits lateral root initiation and primary root elongation [39] while being essential for crown root development in O. sativa [4]. According to the enrichment analysis results, predicted proteins HK4, HK6, and ORR5 are in that pathway, and their network neighbors (CR4, Os11t0143300-02, Os04t0445300-01) may also have a role in root development.

Results of enrichment analysis provide strong evidence to conclude that several predicted proteins are involved in root development, and it is safe to speculate the predicted proteins as accurate predictions, which validates the prediction method.

Identification and functional analysis of sub-modules

There were 6 identified sub-modules (Fig. 3) in the root development PPI network module. Enriched ontology terms were used to describe each sub-module. Proteins and enriched GO-BP terms for each sub-module are given below.

Sub-module (1)

Tables 3 and 4 contain the proteins and the enriched GO-BPs for sub-module (1). According to the results of the enrichment analysis in Table 4, it can be speculated that sub-module (1) is involved with the endoplasmic reticulum (ER) tubular network organization (GO:0071786). The ER tubular network is involved with root hair tip growth [40]. Though the hub protein RHD3 was not annotated to the ER tubular network organization (Table 4), it is involved with root hair tip growth by organizing the ER tubular network [40, 41]. Also, Arabidopsis rhd3 mutant causes short and wavy root hairs [42]. Furthermore, HVA22d protein co-localizes with RHD3 and involves with ER shaping [41], and several proteins of the HVA22 family (protein HVA22-like protein, HVA22-like protein k, HVA22-like protein f, HVA22-like protein e, and HVA22-like protein a), which were predicted, can be seen under this sub-module. According to this information, predicted HVA22 family proteins can be speculated to be involved with root development in O. sativa.

Table 3 NCBI gene descriptions, predicted status, and hub status of the 19 proteins in sub-module (1)
Full size table
Table 4 Significant enriched GO-BP terms from the enrichment analysis for sub-module (1)
Full size table

Proteins Osj_34384, Os12t0604600-01, Os04t0672900-01, and Osj_07614 were also recognized as hubs in this sub-module. Among these, proteins Os04t0672900-01 and Osj_07614 were two predicted candidates which had not been characterized. According to the results, they probably have important roles in root development.

Sub-module (2)

Tables 5 and 6 contain the proteins and the enriched GO-BPs for sub-module (2), respectively.

Table 5 NCBI gene descriptions, predicted status, and hub status of the 34 proteins in sub-module (2)
Full size table
Table 6 Selected enriched GO-BPs and KEGG pathways from the enrichment analysis for sub-module (2)
Full size table

According to Table 6, the most enriched function for the sub-module (2) is the auxin-activated signaling pathway (GO:0009734), and most of the other functions of sub-module (2) are also related to the phytohormone auxin. The auxin signaling pathway combines transport inhibitor response 1 (TIR1), auxin response factors (ARFs), and auxin/indole acetic acid (AUX/IAA) transcriptional repressors together [38]. According to the enrichment analysis results, hub proteins Osj_019622, Osj_14626, PIN1, PIN2, PIN3A, IAA13, and IAA30 and non-hub proteins Osj_28507, Os11t0515500-01, IAA14, ARF12, ARF19, and ARF25 in sub-module (2) participate in the auxin signaling pathway.

Auxin polar transport (GO:0009926) is another enriched function for the sub-module (2) which includes PIN2, PIN1, PIN3A, Osj_28507, and Os08t0564300-04 as annotated proteins. The protein PIN2, a hub with the highest degree of 22, is a potential candidate gene for improving root system architecture in O. sativa [43, 44], and PIN1 and PIN3A are also hubs, which are central to the stability of this sub-module. Auxin polar transport is regulated by the PIN-FORMED (PIN) efflux carriers. PIN polarity plays a crucial role in developing proper organs and proliferation in root proximal meristem [37, 38, 45]. For example, the intramodular hub protein PIN2 works for root development by positioning and emerging root hairs [37], and according to Inahashi et al. [46], the OsPIN2 gene regulates the seminal root elongation and lateral root formation in O. sativa. Moreover, overexpression of the OsPIN2 significantly decreases the number of adventitious roots and the total root length by 22–28% [43]. Furthermore, overexpression of the gene OsPIN3a has led to the development of longer roots and more adventitious roots [47], and Hang et al. [47] suggest that crown root development is controlled by auxin signaling through PIN proteins. Another member of this sub-module, the protein PIN1, is also a hub protein, and overexpression of gene OsPIN1 increases the emergence of adventitious roots, the primary root length, and the number of lateral roots [45, 48].

The roots have the ability to change their growing orientation in response to the changes in gravity [37, 44], and it is controlled by the asymmetric distribution of auxin at the root tip. PIN family, AUX1 (AUXIN-INSENSITIVE1), and other members of the auxin transport pathway contribute to this auxin distribution [37, 44]. Deletion in the OsPIN2 gene has displayed gravitropic root growth phenotypes as shootward auxin distribution in the lower side of the root is largely repressed during a gravity stimulus by the mutation of OsPIN2 [43, 44]. This shows that PIN2 is also essential for root gravitropism. Moreover, the crown root growth angle is an important component of the O. sativa roots, and the OsPIN2 mutant, lra1, has displayed larger root angles [44]. This indicates that the protein PIN2 is important in regulating crown root growth angle. Furthermore, OsPIN3a has shown a notable upregulation in OsPIN2 mutant lra1 since OsPIN3a compensates for the loss of OsPIN2 (agravitropic root phenotype) to some extent [44]. One of the enriched terms for the sub-module (2) is positive gravitropism (GO:0009958), and the seed protein PIN2 and the predicted candidate Os08t0423600-00 have been annotated to that process. In conclusion, this evidence prove that the majority of the sub-module 2 proteins contribute to root development via auxin regulation.

Sub-module (3)

Tables 7 and 8 contain the proteins and the enriched GO-BPs for sub-module (3), respectively. According to Table 8, the top enriched GO-BP term for sub-module (3), cytokinin-activated signaling pathway (GO:0009736), is mediated by a two-component system, and the signaling is transmitted by transcription activators and repressors in a phosphorylation signal transduction system (GO:0000160) [49, 50]. The two-component system comprises three functional modules: sensory histidine kinase (HK), histidine phosphate transfer protein (HP), and response regulator (RR). Cytokinins are sensed by membrane-located HK receptors that transmit signals via HPs to nuclear RRs that activate or repress transcription [51].

Table 7 NCBI gene descriptions, predicted status, and hub status for the 14 proteins in sub-module (3)
Full size table
Table 8 Selected enriched GO-BPs and KEGG pathways of enrichment analysis for sub-module (3)
Full size table

The phytohormone cytokinin is present in the O. sativa root [43] and participates in regulating root development and root architecture. This indicates the involvement of the cytokinin-activated signaling pathway in O. sativa root development and the involvement of sub-module (3) proteins in root development through the cytokinin-activated signaling transduction pathway.

Sub-module (4)

Tables 9 and 10 contain the proteins and the enriched GO-BPs in sub-module (4).

Table 9 NCBI gene descriptions, predicted status, and hub status for the 18 proteins in sub-module (4)
Full size table
Table 10 Selected enriched GO-BPs of enrichment analysis for sub-module (4)
Full size table

As given in Table 10, sub-module (4) has proteins that are associated with both plant cell wall organization (GO:0071555) and cellulose catabolic process (GO:0030245) pathways. The proteins GLU5, GLU3, and Os06t0697000-02 are involved with cell wall organization, and GLU5 and GLU3 are involved with cellulose catabolic process. Cell walls are important in any form of plant development, and cellulose is the major component of the plant cell wall. Cellulose biosynthesis and cell wall loosening enable turgor-driven cell expansion in growing plants, and it has been speculated that endo-1,4-b-glucanases (EGases) play a central role in these complex activities [52]. GLU3 and GLU5 (GLUs) have been directly annotated to the hydrolysis of endo-1,4-b-glucanases [53, 54]. GLUs play important roles in root development, and glu mutants have reduced root development [54]. Furthermore, GLU5 is expressed in lateral root primordia during auxin-induced lateral root development [53].

Phosphate ion transport (GO:0006817) in a plant is mediated by several transporter protein families such as the Pht1 family [55]. PT2 and PT3 belong to the pht1 family and are predicted candidates in sub-module (4). Both root hair length and frequency increase in response to phosphate (Pi) starvation, and the gene expression of OsPT2 is increased during Pi starvation. Therefore, it is reasonable to speculate that PT2, which is a predicted candidate protein, probably has a direct association with Pi transport [55, 56].

FH1 is an intramodular hub in sub-module (4) and also an intermodular hub. FH1 has been directly annotated to root hair development [57, 58]. EXPA8 in sub-module 4 (degree = 7) was not considered a hub according to the hub selection criteria of this study. However, it is a root-specific expansin protein, and expansins are plant cell wall proteins that are involved in cell wall modifications [59, 60]. Overexpression of the OsEXPA8 gene has shown improved root system architecture with longer primary roots and a higher number of lateral roots and root hairs [59, 60]. Moreover, repression of OsEXPA8 has reduced the cell size of the root vascular system and plant height [60]. This evidence proves that sub-module (4) is linked to root development via cell wall organization.

Proteins of other sub-modules are listed in Supplementary Table S4. They did not have any significant enriched GO-BP terms related to root development and need further investigation to confirm their involvement in root development.

Identification and analysis of hub proteins

There are two types of hubs: intramodular and intermodular hubs [22]. Intramodular hubs are found within a functional module, while intermodular hubs act between the modules to interconnect them [21,22,23].

Intramodular hubs

For this study, the top 10% of proteins with the highest degrees were selected [61] as intramodular hubs, which corresponds to a degree cutoff of 8, resulting in 20 proteins (Table 11).

Table 11 Details of intramodular hub proteins
Full size table

Some of the hubs are from the seed proteins, and others are predicted candidates for root development in O. sativa. The predicted hubs are important findings because their relevance in root development has not been revealed to date. They also confirm the accuracy and importance of the predictions. Among the predicted hubs, there were candidates annotated to the enriched GO-BP terms related to the root development in O. sativa, although they lack direct experimental evidence for root development. For example, Os04t0672900-01 and Osj_07614 of sub-module (1) were annotated to GO:0071786: endoplasmic reticulum tubular network organization (Table 4), which is a GO-BP term associated with root development.

Moreover, transcription factors LAX1 (LAX PANICLE 1) and GNP4 (LAX PANICLE 2) are predicted candidates, which were identified as hubs. In O. sativa, LAX1 and GNP4 are required for the formation of axillary meristem throughout the plant’s lifespan [62, 63]. Also, LAX1 shows non-cell-autonomous action (mutant extends beyond the mutant cells); however, its molecular basis has not been revealed yet [62]. Although the functions of lax genes in O. sativa panicle have been studied, their functions in the root are yet to be revealed. The above results provide evidence for their involvement in root development. Therefore, this study provides potential candidates for selecting important proteins for future O. sativa root development studies.

The seed protein OSEIL1 is an intramodular hub and is involved in root development in O. sativa. It is a transcription factor participating in the ethylene signaling pathway, which promotes O. sativa root elongation [64]. Most importantly, OSEIL1 connects with 8 predicted candidates and joins the sub-modules (4) and (6) together (Fig. 3). Therefore, according to our results, OSEIL1 is a critical protein for root development and a likely candidate for future genetic engineering studies.

Intermodular hubs

Intermodular hubs connect different sub-modules (Fig. 3) and are important in linking the different metabolic/biological pathways. Two intermodular hubs were detected in this study (Table 12).

Table 12 Details of intermodular hub proteins
Full size table

As shown in Table 12, the protein DRO1 (DEEPER ROOTING 1) in sub-module 4 works as an intermodular hub and connects sub-modules 2, 3, and 4. Therefore, disturbance to the DRO1 can potentially disrupt the interconnectivity of the pathways or the proper mechanism of those sub-modules. Analysis of DRO1 using iDNA-Prot (http://www.jci-bioinfo.cn/iDNA-Prot) [65], which is a web tool for identifying DNA binding domains in proteins [65], revealed that it may be a DNA-binding protein. Highly expressed DRO1 gene is involved in the regulation of deep rooting by increasing root growth angle which promotes the root growth in a more downward direction [66]. Furthermore, DRO1 enhances nitrogen uptake and cytokinin fluxes at late stages of development by deep rooting which resulted in a high yield in O. sativa. Therefore, DRO1 can be used to develop O. sativa cultivars that have high yields under both drought and non-drought conditions by controlling the root system architecture [66, 67]. As shown in this study, the intermodular hub DRO1 plays a major role in interconnecting and potentially regulating the 3 submodules of root development and can be a valuable candidate for further experimental studies.

The FH1 (formin-like protein 1), which is in the sub-module (4), is an intramodular hub and a critical regulator of the O. sativa root hair development. Osfh1 mutant exhibited growth defects on root hairs. These defects depend on the environmental conditions and were only exhibited when roots were submerged in a solution [57]. According to Huang et al. [58], the external supplies of nutrients or hormones could not rescue the defective mutant. Therefore, FH1 is a crucial protein for the growth of O. sativa as rice is grown under water-logged conditions in the field until the fruit ripening stage [57, 58]. FH1 is also identified as an intermodular hub, and it connects sub-modules (1), (3), and (4) (Table 12, Fig. 3). Formins regulate the growth and elongation of the root hairs and cell wall loosening and synthesis, which are required for root hair development [57]. Furthermore, the sub-module (4) which is annotated to the cell wall organization and the sub-module (1) which is mainly recognized for root hair development are connected by FH1. Since it is connected with 3 sub-modules, it could have more roles in different pathways which are not yet revealed.

Discussion

The large number of proteins interacting with each other during the development of various plant systems makes disentangling their roles using traditional experimental techniques a daunting task. Computational approaches are ideal here as they enable the aggregation of data to construct higher-level views of biological systems (such as PPI network graphs), which are much better at identifying complex relationships between proteins.

During this study, we predicted 75 novel protein candidates associated with root development. Validation of these predictions and analysis of hubs and sub-modules justified that some predictions are annotated to biological processes associated with root development, which confirmed the accuracy of the predictions. These predictions are based on network-based candidate gene prediction, which has been proven to be an accurate method [15]. We also predicted 20 intramodular hubs and 2 intermodular hubs, which are important outcomes of this study. This enabled identifying the important module members central to the stability of the root development in O. sativa. These hub proteins are potential candidates for future genetic engineering experiments as their influence on root development is larger than other proteins because of their centralized nature. The association of several identified hub proteins, such as PIN2 and DRO1, with root development is already experimentally validated. However, there are predicted hub proteins, such as LAX1 and GNP4, which require further experimental investigation. Therefore, this study provides a plethora of protein candidates for future experiments.

Importantly, this study depicts the organization of PPI interactions underlying root development. We unraveled how proteins associated with root development are organized into 6 major sub-modules, mainly attributed to biological processes, such as ER tubular network organization, auxin regulation pathway, cytokinin signaling pathway, and cell wall synthesis. The knowledge about molecular mechanisms and properties of a majority of these module proteins are still incomplete; hence, this analysis provides clues to their collective role in regulating root development in O. sativa.

Conclusion

In this study, we analyzed the network structure of root development proteins, during which 75 new protein candidates, 6 sub-modules, 20 intramodular hubs, and 2 intermodular hubs were identified using a computational analysis. This opens up new directions for future wet lab and dry lab studies based on predicted candidates and hub proteins. To our knowledge, this is the first study that analyzes the PPI network module for root development in O. sativa. Therefore, these findings are the first to show the PPI interaction structure underlying root development, which depicts the importance and applicability of network analysis on other plant developmental phenotypes as well.

Availability of data and materials

All data generated or analyzed during this study are included in this published article and in its accompanying supplementary information files. Python scripts written for this analysis are available at https://github.com/Samadhi9/PPIN-analysis.

Abbreviations

PPI:

Protein-protein interactions

IAA:

Indole-3-acetic acid

NAA:

Naphthaleneacetic acid

IDs:

Identifiers

GO:

Gene Ontology

GO-BP:

Biological process component of the Gene Ontology

ER:

Endoplasmic reticulum

PIN:

PIN-FORMED

HK:

Histidine kinase

HP:

Histidine phosphor transfer

RR:

Response regulator

Pi:

Phosphate

FH2:

Formin homology-2

PR:

Primary root

CR:

Crown root

ARF:

Auxin response factor

EGases:

Endo-1,4-b-glucanase

References

  1. Dixit S, Singh A, Kumar A (2014) Rice breeding for high grain yield under drought: a strategic solution to a complex problem. Int J Agron 2014:1–5

    Article  Google Scholar 

  2. Khan S, Javed MA, Jahan N, Manan FA (2016) A short review on the development of salt tolerant cultivars in rice. Int J Public Health Sci 5:201–212

    Google Scholar 

  3. Rabello AR, Guimarães CM, Rangel PH, da Silva FR, Seixas D, de Souza E, Brasileiro AC, Spehar CR, Ferreira ME, Mehta  (2008) Identification of drought-responsive genes in roots of upland rice (Oryza sativa L). BMC Genomics 9:485. https://doi.org/10.1186/1471-2164-9-485

    Article  Google Scholar 

  4. Meng F, Xiang D, Zhu J, Li Y, Mao C (2019) Molecular mechanisms of root development in rice. Rice 12:1. https://doi.org/10.1186/s12284-018-0262-x

    Article  Google Scholar 

  5. Shoemaker BA, Panchenko AR (2007) Deciphering protein–protein interactions. Part I. Experimental techniques and databases. PLoS Comput Biol 3:e42

    Article  Google Scholar 

  6. Spirin V, Mirny LA (2003) Protein complexes and functional modules in molecular networks. Proc Natl Acad Sci 100:12123–12128

    Article  Google Scholar 

  7. Szklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M, et al. (2017) The STRING database in 2017: quality-controlled protein–protein association networks, made broadly accessible. Nucleic Acids Res 45:D362–8. https://doi.org/10.1093/nar/gkw937.

  8. Wang X, Gulbahce N, Yu H (2011) Network-based methods for human disease gene prediction. Brief Funct Genomics 10:280–293. https://doi.org/10.1093/bfgp/elr024

    Article  Google Scholar 

  9. Franceschini A, Szklarczyk D, Frankild S, Kuhn M, Simonovic M, Roth A, Lin J, Minguez P, Bork P, Von Mering C (2012) STRING v9. 1: protein-protein interaction networks, with increased coverage and integration. Nucleic Acids Res 41:D808–D815

    Article  Google Scholar 

  10. Von Mering C, Jensen LJ, Snel B, Hooper SD, Krupp M, Foglierini M, Jouffre N, Huynen MA, Bork P (2005) STRING: known and predicted protein–protein associations, integrated and transferred across organisms. Nucleic Acids Res 33:D433–D437

    Article  Google Scholar 

  11. Fernando SPC (2018) Investigating evolutionary phenotypic transitions using integrative biological network frameworks. University of South Dakota. https://www.proquest.com/openview/8dc3742ab689ea9eea93ee7a14cace6d/1?pq-10.1186/s43141-023-00515-8

  12. Fernando PC, Mabee PM, Zeng E (2020) Integration of anatomy ontology data with protein–protein interaction networks improves the candidate gene prediction accuracy for anatomical entities. BMC Bioinformatics 21:442. https://doi.org/10.1186/s12859-020-03773-2

    Article  Google Scholar 

  13. Ekman D, Light S, Björklund ÅK, Elofsson A (2006) What properties characterize the hub proteins of the protein-protein interaction network of Saccharomyces cerevisiae? Genome Biol 7:R45. https://doi.org/10.1186/gb-2006-7-6-r45

    Article  Google Scholar 

  14. Hishigaki H, Nakai K, Ono T, Tanigami A, Takagi T (2001) Assessment of prediction accuracy of protein function from protein–protein interaction data. Yeast 18:523–531

    Article  Google Scholar 

  15. Sharan R, Ulitsky I, Shamir R (2007) Network-based prediction of protein function. Mol Syst Biol 3:88

    Article  Google Scholar 

  16. Saini A, Hou J (2013) Progressive clustering based method for protein function prediction. Bull Math Biol 75:331–350

    Article  MathSciNet  MATH  Google Scholar 

  17. Deng M, Zhang K, Mehta S, Chen T, Sun F. Prediction of Protein Function Using Protein–Protein Interaction Data. J Comput Biol 2003;10:947–60. https://doi.org/10.1089/106652703322756168

  18. Park D, Singh R, Baym M, Liao C-S, Berger B (2010) IsoBase: a database of functionally related proteins across PPI networks. Nucleic Acids Res 39:D295–D300

    Article  Google Scholar 

  19. Lan W, Wang J, Li M, Peng W, Wu F (2015) Computational approaches for prioritizing candidate disease genes based on PPI networks. Tsinghua Sci Technol 20:500–512

    Article  MathSciNet  Google Scholar 

  20. Park S-J, Choi J-S, Kim B-C, Jho S-W, Ryu J-W, Park D, et al. PutidaNET: Interactome database service and network analysis of Pseudomonas putida KT2440. BMC Genomics 2009;10:S18. https://doi.org/10.1186/1471-2164-10-S3-S18

  21. Kiran M, Nagarajaram HA (2016) Interaction and localization diversities of global and local hubs in human protein–protein interaction networks. Mol BioSyst 12:2875–2882

    Article  Google Scholar 

  22. Han J-DJ, Bertin N, Hao T, Goldberg DS, Berriz GF, Zhang LV, Dupuy D, Walhout AJ, Cusick ME, Roth FP (2004) Evidence for dynamically organized modularity in the yeast protein–protein interaction network. Nature 430:88–93

    Article  Google Scholar 

  23. Chang X, Xu T, Li Y, Wang K (2013) Dynamic modular architecture of protein-protein interaction networks beyond the dichotomy of ‘date’and ‘party’hubs. Sci Rep 3:1691

    Article  Google Scholar 

  24. Liu J, Jing L, Tu X (2016) Weighted gene co-expression network analysis identifies specific modules and hub genes related to coronary artery disease. BMC Cardiovasc Disord 16:54

    Article  Google Scholar 

  25. Apic G, Ignjatovic T, Boyer S, Russell RB (2005) Illuminating drug discovery with biological pathways. FEBS Lett 579:1872–1877. https://doi.org/10.1016/j.febslet.2005.02.023

    Article  Google Scholar 

  26. Liseron-Monfils C, Ware D (2015) Revealing gene regulation and associations through biological networks. Curr Plant Biol 3–4:30–39. https://doi.org/10.1016/j.cpb.2015.11.001

    Article  Google Scholar 

  27. Albert R (2007) Network inference, analysis, and modeling in systems biology. Plant Cell 19:3327–3338. https://doi.org/10.1105/tpc.107.054700

    Article  Google Scholar 

  28. Winterbach W, Mieghem PV, Reinders M, Wang H, de Ridder D (2013) Topology of molecular interaction networks. BMC Syst Biol 7:90. https://doi.org/10.1186/1752-0509-7-90

    Article  Google Scholar 

  29. Gustafsson M, Nestor CE, Zhang H, Barabási A-L, Baranzini S, Brunak S, Chung KF, Federoff HJ, Gavin A-C, Meehan RR, Picotti P, Pujana MÁ, Rajewsky N, Smith KG, Sterk PJ, Villoslada P, Benson M (2014) Modules, networks and systems medicine for understanding disease and aiding diagnosis. Genome Med 6:82. https://doi.org/10.1186/s13073-014-0082-6

    Article  Google Scholar 

  30. Liu W, Wu A, Pellegrini M, Wang X (2015) Integrative analysis of human protein, function and disease networks. Sci Rep 5:14344. https://doi.org/10.1038/srep14344

    Article  Google Scholar 

  31. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T (2003) Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13:2498–2504

    Article  Google Scholar 

  32. Smoot ME, Ono K, Ruscheinski J, Wang P-L, Ideker T (2011) Cytoscape 2.8: new features for data integration and network visualization. Bioinformatics 27:431–432

    Article  Google Scholar 

  33. Huang DW, Sherman BT, Lempicki RA (2009) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 4(1):44–57

    Article  Google Scholar 

  34. Morris JH, Apeltsin L, Newman AM, Baumbach J, Wittkop T, Su G, Bader GD, Ferrin TE (2011) clusterMaker: a multi-algorithm clustering plugin for Cytoscape. BMC Bioinformatics 12:436

    Article  Google Scholar 

  35. Fan W, Ye G (2018) Microarray analysis for the identification of specific proteins and functional modules involved in the process of hepatocellular carcinoma originating from cirrhotic liver. Mol Med Rep 17:5619–5626

    Google Scholar 

  36. Wang H, Niu Q-W, Wu H-W, Liu J, Ye J, Yu N, Chua N-H (2015) Analysis of non-coding transcriptome in rice and maize uncovers roles of conserved lnc RNA s associated with agriculture traits. Plant J 84:404–416

    Article  Google Scholar 

  37. Leyser O (2018) Auxin signaling. Plant Physiol 176:465–479

    Article  Google Scholar 

  38. Takatsuka H, Umeda M (2014) Hormonal control of cell division and elongation along differentiation trajectories in roots. J Exp Bot 65:2633–2643

    Article  Google Scholar 

  39. Kieber JJ, Schaller GE (2018) Cytokinin signaling in plant development. Development 145:dev149344

    Article  Google Scholar 

  40. Zheng H, Chen J (2011) Emerging aspects of ER organization in root hair tip growth: lessons from RHD3 and Atlastin. Plant Signal Behav 6:1710–1713

    Article  Google Scholar 

  41. Chen J, Stefano G, Brandizzi F, Zheng H (2011) Arabidopsis RHD3 mediates the generation of the tubular ER network and is required for Golgi distribution and motility in plant cells. J Cell Sci 124:2241–2252

    Article  Google Scholar 

  42. Wang H, Lockwood SK, Hoeltzel MF, Schiefelbein JW (1997) The ROOT HAIR DEFECTIVE3 gene encodes an evolutionarily conserved protein with GTP-binding motifs and is required for regulated cell enlargement in Arabidopsis. Genes Dev 11:799–811

    Article  Google Scholar 

  43. Chen Y, Fan X, Song W, Zhang Y, Xu G (2012) Over-expression of OsPIN2 leads to increased tiller numbers, angle and shorter plant height through suppression of OsLAZY1. Plant Biotechnol J 10:139–149

    Article  Google Scholar 

  44. Wang L, Guo M, Li Y, Ruan W, Mo X, Wu Z, Sturrock CJ, Yu H, Lu C, Peng J (2018) LARGE ROOT ANGLE1, encoding OsPIN2, is involved in root system architecture in rice. J Exp Bot 69:385–397

    Article  Google Scholar 

  45. Miyashita Y, Takasugi T, Ito Y (2010) Identification and expression analysis of PIN genes in rice. Plant Sci 178:424–428

    Article  Google Scholar 

  46. Inahashi H, Shelley IJ, Yamauchi T, Nishiuchi S, Takahashi-Nosaka M, Matsunami M, Ogawa A, Noda Y, Inukai Y (2018) OsPIN2, which encodes a member of the auxin efflux carrier proteins, is involved in root elongation growth and lateral root formation patterns via the regulation of auxin distribution in rice. Physiol Plant 164:216–225

    Article  Google Scholar 

  47. Zhang Q, Li J, Zhang W, Yan S, Wang R, Zhao J, Li Y, Qi Z, Sun Z, Zhu Z (2012) The putative auxin efflux carrier OsPIN3t is involved in the drought stress response and drought tolerance. Plant J 72:805–816

    Article  Google Scholar 

  48. Xu M, Zhu L, Shou H, Wu P (2005) A PIN1 family gene, OsPIN1, involved in auxin-dependent adventitious root emergence and tillering in rice. Plant Cell Physiol 46:1674–1681

    Article  Google Scholar 

  49. Heyl A, Schmülling T (2003) Cytokinin signal perception and transduction. Curr Opin Plant Biol 6:480–488

    Article  Google Scholar 

  50. Müller B, Sheen J (2007) Advances in cytokinin signaling. Science 318:68–69

    Article  Google Scholar 

  51. Du L, Jiao F, Chu J, Jin G, Chen M, Wu P (2007) The two-component signal system in rice (Oryza sativa L.): a genome-wide study of cytokinin signal perception and transduction. Genomics 89:697–707

    Article  Google Scholar 

  52. Levy I, Shani Z, Shoseyov O (2002) Modification of polysaccharides and plant cell wall by endo-1, 4-β-glucanase and cellulose-binding domains. Biomol Eng 19:17–30

    Article  Google Scholar 

  53. Yoshida K, Imaizumi N, Kaneko S, Kawagoe Y, Tagiri A, Tanaka H, Nishitani K, Komae K (2006) Carbohydrate-binding module of a rice endo-β-1, 4-glycanase, OsCel9A, expressed in auxin-induced lateral root primordia, is post-translationally truncated. Plant Cell Physiol 47:1555–1571

    Article  Google Scholar 

  54. Zhou H-L, He S-J, Cao Y-R, Chen T, Du B-X, Chu C-C, Zhang J-S, Chen S-Y (2006) OsGLU1, a putative membrane-bound endo-1, 4-ß-d-glucanase from rice, affects plant internode elongation. Plant Mol Biol 60:137–151

    Article  Google Scholar 

  55. Smith FW, Mudge SR, Rae AL, Glassop D (2003) Phosphate transport in plants. Plant Soil 248:71–83

    Article  Google Scholar 

  56. Paszkowski U, Kroken S, Roux C, Briggs SP (2002) Rice phosphate transporters include an evolutionarily divergent gene specifically activated in arbuscular mycorrhizal symbiosis. Proc Natl Acad Sci 99:13324–13329

    Article  Google Scholar 

  57. Huang J, Kim CM, Xuan Y, Liu J, Kim TH, Kim B-K, Han C (2013) Formin homology 1 (OsFH1) regulates root-hair elongation in rice (Oryza sativa). Planta 237:1227–1239

    Article  Google Scholar 

  58. Huang J, Liu J, Han C (2013) Formin homology 1 (OsFH1) regulates submergence-dependent root hair development in rice plants. Plant Signal Behav 8:1227–1239

    Article  Google Scholar 

  59. Ma N, Wang Y, Qiu S, Kang Z, Che S, Wang G, Huang J (2013) Overexpression of OsEXPA8, a root-specific gene, improves rice growth and root system architecture by facilitating cell extension. PLoS One 8(10):e75997

    Article  Google Scholar 

  60. Wang Y, Ma N, Qiu S, Zou H, Zang G, Kang Z, Wang G, Huang J (2014) Regulation of the α-expansin gene OsEXPA8 expression affects root system architecture in transgenic rice plants. Mol Breed 34:47–57

    Article  Google Scholar 

  61. Flórez AF, Park D, Bhak J, Kim B-C, Kuchinsky A, Morris JH, Espinosa J, Muskus C (2010) Protein network prediction and topological analysis in Leishmania major as a tool for drug target selection. BMC Bioinformatics 11:484. https://doi.org/10.1186/1471-2105-11-484

    Article  Google Scholar 

  62. Oikawa T, Kyozuka J (2009) Two-step regulation of LAX PANICLE1 protein accumulation in axillary meristem formation in rice. Plant Cell 21:1095–1108

    Article  Google Scholar 

  63. Tabuchi H, Zhang Y, Hattori S, Omae M, Shimizu-Sato S, Oikawa T, Qian Q, Nishimura M, Kitano H, Xie H (2011) LAX PANICLE2 of rice encodes a novel nuclear protein and regulates the formation of axillary meristems. Plant Cell 23:3276–3287

    Article  Google Scholar 

  64. Mao C, Wang S, Jia Q, Wu P (2006) OsEIL1, a rice homolog of the Arabidopsis EIN3 regulates the ethylene response as a positive component. Plant Mol Biol 61:141

    Article  Google Scholar 

  65. Lin W-Z, Fang J-A, Xiao X, Chou K-C (2011) iDNA-Prot: identification of DNA binding proteins using random forest with grey model. PLoS One 6(9):e24756

    Article  Google Scholar 

  66. Uga Y, Sugimoto K, Ogawa S, Rane J, Ishitani M, Hara N, Kitomi Y, Inukai Y, Ono K, Kanno N (2013) Control of root system architecture by DEEPER ROOTING 1 increases rice yield under drought conditions. Nat Genet 45:1097

    Article  Google Scholar 

  67. Arai-Sanoh Y, Takai  T, Yoshinaga S, Nakano H, Kojima M, Sakakibara H, Kondo M, Uga Y (2014) Deep rooting conferred by DEEPER ROOTING 1 enhances rice yield in paddy fields Scientific Reports 4(1). https://doi.org/10.1038/srep05563

Download references

Funding

This research was supported by the University of Colombo funds for undergraduate research.

Author information

Authors and Affiliations

  1. Department of Plant Sciences, Faculty of Science, University of Colombo, Colombo, Sri Lanka

    Samadhi S. Wimalagunasekara, Janith W. J. K. Weeraman, Shamala Tirimanne & Pasan C. Fernando

Authors
  1. Samadhi S. Wimalagunasekara
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Janith W. J. K. Weeraman
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Shamala Tirimanne
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Pasan C. Fernando
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

All authors planned and designed the experiments. SW wrote the Python scripts for the analysis and performed the experiments under the supervision of ST and PF. JW assisted in preparing the final manuscript. All authors analyzed the results and read and approved the final manuscript.

Corresponding author

Correspondence to Janith W. J. K. Weeraman.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplementary Table S1.

Details of seed proteins. Supplementary Table S2. List of NCBI gene symbols of proteins of extracted network module and the prediction scores for the top 75 candidates. Supplementary Table S3. Details of missing seed proteins. Supplementary Table S4. Details of sub-modules.

Additional file 2: Supplementary Fig. 1.

PPI network module visualization with 100 predicted proteins and 45 seed proteins.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wimalagunasekara, S.S., Weeraman, J.W.J.K., Tirimanne, S. et al. Protein-protein interaction (PPI) network analysis reveals important hub proteins and sub-network modules for root development in rice (Oryza sativa). J Genet Eng Biotechnol 21, 69 (2023). https://doi.org/10.1186/s43141-023-00515-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s43141-023-00515-8

Keywords