Maize inbred lines, Mp708 and Tx601, used in this study were grown in soil mixed with vermiculite and perlite (PRO-MIX BX BIOFUNGICIDE + MYCORRHIZAE, Premier Tech Horticulture Ltd., Canada) in growth chambers with 14:10 (L:D) h photoperiod, 160 μE light m^–2 s^–1, 25°C, and 50–60% relative humidity. All plants for the experiments were used at the V2-V3 developmental stage (∼2 weeks; Ritchie et al., 1998). These plants were grown in 3.8 cm × 21.0 cm plastic Cone-tainers (Hummert International, Earth City, MO, United States). ECB and WCR were obtained from Benzon Research Inc., PA, and Crop Characteristics Inc., MN, respectively. ECB eggs obtained from Benzon Research Inc. were reared on artificial diet in growth chamber with 14:10 (L:D) h photoperiod at 23°C as described previously (Louis et al., 2013). Non-diapausing WCR eggs were supplied by Crop Characteristics Inc. in 2015. WCRs were originally collected from susceptible non-Bt maize fields in Minnesota, United States, and this colony was maintained on non-Bt maize plants at Crop Characteristics Inc. for more than 80 generations (Scott Sargent, Crop Characteristics Inc. personal communication). Thus, these WCR populations were never exposed to selection for Bt resistance, and therefore, remain susceptible to Bt plants. WCR eggs obtained were kept in growth chamber with 14:10 (L:D) h photoperiod at 23°C. Larvae were hatched at 23°C in growth chambers and were used for experiments.

For insect bioassays, two newly molted 2nd instar ECB each weighing approximately 8–12 mg were introduced into the whorl region of maize plants. Individual plants were caged and after 5 days ECB larvae were collected. Only one ECB larva was able to be recovered from each maize plant due to ECB cannibalism, and the final weight individual larva was recorded. Similarly, to understand the interaction between ECB and WCR, two newly molted 5th instar ECB were introduced into the whorl region of Mp708 plants. After 24 h, ECB larvae were removed from the whorl region, and two newly hatched neonates of WCR were used to infest each plant. A fine-tipped paint brush (size O) was used to transfer neonate WCR larvae to maize plants as described previously (Jakka et al., 2016). Briefly, newly hatched WCR larvae were placed on the exposed nodal roots of maize plants and a thin layer of soil was used to cover the exposed roots. Bioassay plants were kept in a growth chamber with 14:10 (L:D) h photoperiod, 25°C, and 50–60% relative humidity. These plants were watered as needed. After 12 days, the soil content of Cone-tainers were combined and the WCR were collected as described previously (Gill et al., 2011). Briefly, the soil content of all pants within a treatment was combined and sifted by hand until the collection of WCR was occasional. Subsequently, the soil was sifted through two mesh screens of different sizes (2.000 and 0.833 mm) to find the remaining WCR larvae in the soil. Finally, we ceased to monitor the presence of WCR in the soil content until none were obtained after 15–20 min of incessant searching. Mp708 plants that were uninfested with ECB on the whorl region were used as the control. Both bioassays were replicated twice with similar results and data from both independent experiments were combined and presented in Figure 1. The insect bioassay data were analyzed using mixed-model analysis of variance with experiment replication as the random effect. Analysis of variance (ANOVA) was performed using PROC GLIMMIX (SAS 9.3, SAS Institute, Cary, NC, United States), and means, when significant, were separated using the least significant difference procedure (P < 0.05).

The hallmark of Mp708’s defense is the accumulation of Mir1-CP protein both at the site of insect feeding or in systemic leaves or tissues away from the site of insect feeding (i.e., distal roots; Pechan et al., 2000; Louis et al., 2015; Varsani et al., 2016). To determine the presence of the Mir1-CP protein in Mp708 roots after aboveground feeding by ECB, we performed a western-blot analysis. Protein extraction and western-blot analysis were performed as described previously (Ankala et al., 2009; Varsani et al., 2016). Protein concentration was determined with Qubit reagent (Q33211) and Qubit 4 Fluorometer. A total of 60 μg of protein was used for electrophoresis. Upon gel electrophoresis, proteins were transferred to PVDF membrane using the iBlot 2 gel transfer system. Images were finally captured with iBright FL1000 Imaging System.

At the V2-V3 stage of Mp708 plants, two newly molted 5th instar ECB were introduced into the whorl region. Individual plants were caged, and root samples were collected 24 h after infestation. ECB uninfested Mp708 plants were used as the control. In total, three biological replicates were produced for each treatment. Each biological replicate consisted of three individual maize tissues (root) pooled together. For each replicate, three ECB-infested and uninfested root tissues were collected and flash-frozen in liquid nitrogen. Maize Mp708 root tissues (80–100 mg) were ground using 2010 Geno/Grinder (SPEX SamplePrep, NJ, United States) for 40 s at 1,400 strokes min^–1. Total RNA was extracted from the homogenized tissue using Qiagen RNeasy Plant Mini Kit. Extracted total RNA was quantified through Nanodrop 2000c Spectrophotometer (Thermo Scientific TM). Then, RNA-seq libraries were constructed by using the mRNA-seq standard TruSeq protocol from Illumina. RNA-seq libraries were sequenced in 50 bp paired-end with 20 million reads on average per library.

The quality check of the RNA-seq libraries was performed with FASTQC (Andrews, 2010) and reads with a Phred score lower than 20 and length below 45 base pairs were removed with Trimmomatic v0.39 (Bolger et al., 2014). Then, trimmed reads were mapped on the maize reference genome v4¹ with Tophat2 (Kim et al., 2013) using the following parameters: 1 mismatch (-N 1), 0 splicing mismatch (-m 0), unique mapped reads (-g 1 -M). The transcripts reconstruction was performed with Cufflinks v2.2.1 with the following parameters: quantification against the reference annotation only (-G), multi-read-correct (-u) and frag-biais-correct (-b). The differentially expressed gene analysis was performed with Cuffdiff 2.2.1. Differentially expressed genes (DEGs) were identified with the following parameters: multiple-testing corrected P-values ≤ 5% and expression ratios of | log₂(Infested/Control)| ≥ log₂(2). Gene ontology (GO) were analyzed with MaizeMine² by using the reference annotation as a template.

The GOBU package was used for enrichment calculations (Lin et al., 2006). The full set of maize gene annotation was used as the reference comparison set against down or upregulated DEGs. The P-values were calculated using Fisher’s exact test and corrected for multiple testing with FDR method using the R module called “P-adjust.”

Previously, it was shown that Mp708 provides enhanced resistance to different feeding guilds of insect pests (Pechan et al., 2000; Lopez et al., 2007; Gill et al., 2011; Louis et al., 2015; Varsani et al., 2016; Pingault et al., 2021). Here, we examined whether Mp708 provides enhanced resistance to stem boring insect, ECB. Two independent experiments with similar experimental set up were performed. In the first and second bioassays,10 and 14 plants each of Tx601 and Mp708, respectively, were used and the combined data is presented in Figure 1A. Our results show that ECB growth was lower on the Mp708 plants compared to the maize inbred line Tx601 (Figure 1A). As mentioned before, the hallmark of Mp708’s defense is the accumulation of Mir1-CP at the site (i.e., leaf) and distal parts (i.e., roots) of the plant after insect infestation (Louis et al., 2015; Varsani et al., 2016). To determine if Mir1-CP accumulates in the roots after foliar feeding by ECB, we monitored the Mir1-CP levels in the roots of Mp708 plants using immunoblot analysis. Our results suggest an increased accumulation of Mir1-CP levels in the roots after ECB feeding in the aboveground parts of the plant for 24 h (Supplementary Figure 1). Next, we monitored whether aboveground feeding by ECB on Mp708 plants can induce resistance to subsequent herbivory by belowground feeding of WCR. In the first bioassay, out of 16 WCR larvae introduced on eight Mp708 plants, we recovered 12 and 13 WCR larvae on Mp708 plants that were preinfested with ECB (+ECB) and Mp708 plants that were uninfested with ECB (-ECB), respectively. Similarly, in the second bioassay, out of 24 WCR larvae introduced on 12 Mp708 plants, we recovered 17 and 19 WCR larvae on Mp708 plants that were preinfested with ECB (+ECB) and Mp708 plants that were uninfested with ECB (-ECB), respectively, and the fresh weight of each WCR was recorded. Here, the datasets from two independent experiments with similar experimental set up were combined pooled and the combined data is presented in Figure 1B. Our insect bioassay results indicate that aboveground feeding by ECB for 24 h provided enhanced resistance to subsequent herbivory by belowground feeding of WCR (Figure 1B). Collectively, these data suggest that aboveground feeding by ECB can trigger belowground defenses in maize.

After 24 h of ECB feeding on the whorl region of the Mp708 plants, we identified 4,793 genes that were differentially expressed in Mp708 roots. Among these, 1,629 genes were upregulated, while 3,164 genes were downregulated (Figure 2 and Supplementary Table 1). The gene ontology enrichment was performed for each up and downregulated gene sets. The top ten gene ontology molecular functions terms enriched in upregulated DEGs were related to “protein kinase activity” (141 genes), “carbohydrate binding” (48 genes), “protein serine/threonine kinase activity” (97 genes), “Oxidoreductase activity, acting on paired donors, with incorporation, or reduction of molecular oxygen” (51 genes; Figure 2 and Supplementary Table 2). Molecular function terms enriched in downregulated genes included “peroxidase activity” (69 genes), “oxidoreductase activity” (294 genes), “antioxidant activity” (72 genes; Figure 2 and Supplementary Table 2).

The gene ontology biological process enrichment analysis revealed that downregulated genes are related with functions enriched in “response to oxidative stress,” “oxidation-reduction process,” or “carbohydrate metabolic process” while upregulated genes are related to functions as: “protein phosphorylation,” “regulation of nucleic acid-templated transcription,” “regulation of transcription, DNA-templated” (Supplementary Table 2).

The KEGG pathway analysis revealed three pathways enriched in the upregulated genes: “plant hormones signal transduction” (22 genes), “MAPK signaling pathway” (15 genes) and “Circadian rhythm” (8 genes; Figure 2 and Supplementary Table 2). Downregulated genes were enriched in “biosynthesis of secondary metabolites” (164 genes), “phenylpropanoid biosynthesis” (56 genes), “starch and sucrose metabolism” (33 genes; Figure 2 and Supplementary Table 2). In addition, 10 downregulated genes had function associated with “early nodulin-related” or “early nodulin-like protein.”

Among the DEGs, 670 were transcription factors (TFs) and 18 families were represented by >1% of the TFs (Figure 3). Among the 18 families, 13 were composed of at least 50% of downregulated genes (Figure 3 and Supplementary Table 3), including bHLH (152 DEGs and 95 DEGs downregulated), HB (78 DEGs and 44 DEGs downregulated), NAC (62 DEGs and 32 DEGs downregulated), and AP2 (39 DEGs and 28 DEGs downregulated; Figure 3 and Supplementary Table 3). Conversely, TFs families represented by at least 50% of DEGs upregulated are G2-like (56 DEGs and 36 upregulated DEGs), WRKY (21 DEGs and 12 upregulated DEGs) and ERF (ETHYLENE RESPONSIVE FACTORS, 17 DEGs, and 13 upregulated DEGs; Figure 3 and Supplementary Table 3). The large number of upregulated ERF transcription factors in Mp708 plants further confirm our previous observation that ET acts as a central node in mediating maize defense against various feeding guilds of insect pests (Louis et al., 2015).

To gain a better understanding for the role of phytohormones in defense against root feeders in response to aboveground herbivory, we investigated the DEGs involved in phytohormone-related pathways. Thirty-one DEGs were found and divided between ET (11 DEGs), JA (12 DEGs), and SA (8 DEGs) pathways (Figure 4 and Supplementary Table 4). Among the 11 DEGs related to the ET pathway, six genes were upregulated: Ethylene-insensitive protein 2 (EIN2; Zm00001d039341), Ethylene-Insensitive-Like 8 (EIL8; Zm00001d016924), and four members of the 1-Aminocyclopropane-1-Carboxylic Acid Oxidase (ACO) family (ZmACO6: Zm00001d000163, ZmACO4: Zm00001d024852, ZmACO4: Zm00001d024853, and ZmACO20: Zm00001d052136). Interestingly, two gene members of the Acetyl-CoA synthetase (ACS) family were both downregulated (ACS1: Zm00001d039487 and ACS6: Zm00001d033862; Figure 4 and Supplementary Table 4). In contrast to ET, a majority of the DEGs involved in JA pathway were downregulated (11 genes) while only two genes were upregulated: JASMONATE RESISTANT 1, (JAR1; Zm00001d039345) and OPR5 (Zm00001d003584; Figure 4 and Supplementary Table 4). The three remaining genes of the OPR family, along with the six genes of the OPDA family and two genes of the ACS family, were downregulated (Figure 4 and Supplementary Table 4). All the DEGs involved in the SA pathway were downregulated in the roots of Mp708 plants after ECB feeding on the shoot region for 24 h (Figure 4 and Supplementary Table 4).

Mir1-CP, encoded by mir1, is characterized as a key component for the resistance to insect pests in maize (Pechan et al., 2000; Louis et al., 2015). Our results indicate that mir1 (Zm00001d036542) was upregulated in Mp708 roots at 24 hpi of ECB feeding on the whorl region with a high fold-change (FC; log₂FC = 2.075). Protease inhibitors have also been identified in playing a role in direct plant defense mechanisms in response to insect attack (Koiwa et al., 1997). Here, one gene encoding a serpin, Zm00001d049148, was downregulated. Zm00001d002960, a member of Kunitz type inhibitor (KTI) family, was also downregulated. However, another KTI (smart serine protease inhibitor), WIN2 (Zm00001d015279), was upregulated. Three genes encoding Bowman-Birk type inhibitors were differentially expressed, among them two (Zm00001d003304 and Zm00001d040027) were downregulated and one, Zm00001d012939, was upregulated (Supplementary Table 1). Chitinases have also been identified as playing a role in plant defense and suppressing herbivore defenses (Ray et al., 2016a). Among the DEGs, 11 genes were functionally annotated as “chitinase family related,” “Chinase A,” “basic chitinase,” or “homolog of carrot EP3-3 chitinase.” Among the 11 genes encoding chitinases, five were upregulated (Zm00001d036370, Zm00001d014840, Zm00001d043988, Zm00001d003190, and Zm00001d027525) in the roots of Mp708 plants after ECB feeding for 24 h on the shoot region, with a log₂FC from 1.05 (Zm00001d027525) to 2.38 (Zm00001d036370), and six genes were downregulated (Zm00001d010911, Zm00001d037656, Zm00001d036366, Zm00001d048903, Zm00001d017152, and Zm00001d042769) with a log₂FC from -1.39 to -3.06 (Supplementary Table 1).

Root feeding insects are often attracted to metabolites released from the host plant, such as flavonoids, sugars, amino acids, or terpenes (Erb et al., 2013). Genes involved in these pathways were mined from the maize annotation using KEGG ID (Supplementary Table 5) and the expression profile is provided in Figure 5. One and ten genes were upregulated in the phenylpropanoid and flavonoid pathways, respectively, (Figure 5). Among the upregulated genes, one gene encoded a flavanone 3-hydroxylase (F3H, Zm00001d001960), two genes encoded chalcone and stilbene synthase family proteins (CHS, Zm00001d052673 and Zm00001d007403), and three genes encoded 2-oxoglutarate (2OG) and Fe(II)-dependent oxygenase superfamily proteins (Zm00001d037487, Zm00001d042192, and Zm00001d051896), including DMR6 (DOWNY MILDEW RESISTANT 6, Zm00001d051896). DMR6 was proven to be involved in plant immunity for Arabidopsis (Van Damme et al., 2008). Zm00001d002287 (FLS2, Leucine-rich receptor-like protein kinase family protein) was downregulated with a log₂FC (-6.04). FLS2 was described as a pattern recognition receptor that perceives the PAMP (Chinchilla et al., 2006). Three genes encoding 2-oxoglutarate (2OG) and Fe(II)-dependent oxygenase superfamily proteins were also downregulated (Zm00001d002564, Zm00001d012456, and Zm00001d047744). Three genes encoding UDP-glucosyl transferases were downregulated (Figure 5). Among the DEGs involved in amino acid metabolism, two genes were upregulated (Zm00001d028303, Zm00001d037455), and the remaining 13 genes were downregulated, as well as the seven and five DEGs involved in sugar and terpene pathways, respectively, (Figure 5). These data indicated a complex change in root metabolism within 24 h of aboveground feeding by ECB, with potential changes in the profiles and abundances of metabolites secreted by roots.