**3. Results**

#### *3.1. E*ff*ects of GABA and VGB on Seed Germination Characteristics*

The CK, 10 mM GABA, and 100 μM VGB solutions did not induce seed germination on Day 2. The germination rate of the CK seeds increased after five days, whereas the 10 mM GABA and 100 μM VGB treatments inhibited chestnut seed germination and primary root growth at five days after sowing (Figure 1A). The germination rates of the CK seeds gradually increased by over 27.8–54.8% between Days 5 and 30 (Figure 1B), whereas the germination rates gradually decreased by over 11.0–17.2% and 17.7–21.1% following the GABA and VGB treatments, respectively (Figure 1B). The VGB treatment decreased the average root length per seed between Days 8 and 30 (Figure 1C), whereas the total root length was generally decreased by both GABA and VGB between Days 5 and 30 (*p* < 0.05). These

results indicated that both treatments had significant inhibitory effects on chestnut seed germination and early root growth.

**Figure 1.** (**A**) Effects of exogenous GABA and VGB on (**B**) chestnut seed germination and (**C**,**D**) primary root growth. \* and \*\* represent significant differences between the treatments and the control (CK) at *p* < 0.05 and *p* < 0.01, respectively.

#### *3.2. Changes to Endogenous GABA and H2O2 Contents*

In the untreated seeds, the endogenous GABA concentration was initially 185.5 μg/g FW, but then increased to 767.3 μg/g FW at 48 h, after which it significantly decreased to 422.6 μg/g FW at 120 h (Figure 2A). However, the endogenous GABA concentrations decreased only following the VGB treatment after the 15 h imbibition (0 h). Additionally, the GABA and VGB treatments significantly decreased the endogenous GABA concentrations by over 22.1–23.5% at 48 h. Regarding the H2O2 contents, the CK level slightly increased. Surprisingly, the H2O2 contents increased considerably after the GABA and VGB treatments. After the 15 h imbibition, the H2O2 contents increased sharply by over 94.0% and 163.0% in response to GABA and VGB treatments, respectively, at 0 h, and remained high until 120 h (Figure 2B).

*Plants* **2020**, *9*, 449

**Figure 2.** Effects of exogenous GABA and VGB on endogenous ( **A**) GABA and (**B**) H2O2 contents during chestnut seed germination. \* and \*\* represent significant differences between the treatments and control (CK) at *p* < 0.05 and *p* < 0.01, respectively. t0: time-point when the seed imbibition was initiated.

#### *3.3. E*ff*ects of GABA and VGB on Carbon Metabolism*

Compared with the CK level, the soluble sugar contents significantly increased in response to exogenous GABA only at 48 h, whereas it significantly increased at 0 and 48 h, but decreased at 120 h following the VGB treatment (Figure 3A). However, the GABA and VGB treatments did not significantly affect the starch contents (Figure 3B). Additionally, compared with the corresponding CK levels, both treatments decreased and increased the abundance of the tricarboxylic acid (TCA) cycle intermediates citrate and malate, respectively, but only at 120 h; there were no significant di fferences with the CK levels at all other examined time-points (Figure 4). Specifically, the citrate content decreased by over 34.2% and 65.8% (Figure 4A), and the malate content increased by over 1.8- and 5.1-fold (Figure 4B), in response to the GABA and VGB treatments, respectively. These results indicated that both treatments mainly a ffected the TCA cycle activity after germination.

**Figure 3.** Effects of exogenous GABA and VGB on ( **A**) the soluble sugar and (**B**) starch contents during chestnut seed germination. \* represents a significant di fference between the treatments and control (CK) at *p* < 0.05. t0: time-point when the seed imbibition was initiated.

#### *3.4. E*ff*ects of GABA and VGB on Nitrogen Metabolism*

Relative to the CK levels, the soluble protein contents increased in response to exogenous GABA and VGB at 120 h (Figure 5A). In contrast, only VGB increased the total amino acid content (0 h) (Figure 5B). The abundance of the following 12 free amino acids decreased after the GABA and VGB treatments at the 48 h time-point: Ser, His, Arg, Thr, Pro, Tyr, Val, Met, Lys, Ile, Phe, and Leu. At 120 h, only the Asp content was increased by both treatments, whereas the abundance of 13 free amino acids increased only in response to VGB (i.e., Ser, Glu, Gly, His, Arg, Thr, Pro, Tyr, Val, Lys, Ile, Phe, and Leu) (Figure 6). These results implied that the GABA and VGB treatments negatively regulated amino acid

metabolism before germination, but after germination, the free amino acid contents were increased mainly by VGB.

**Figure 4.** Effects of exogenous GABA and VGB on (**A**) the citrate and (**B**) malate contents during chestnut seed germination. \* represents a significant difference between the treatments and control (CK) at *p* < 0.05. t0: time-point when the seed imbibition was initiated.

**Figure 5.** Effects of exogenous GABA and VGB on (**A**) soluble protein and (**B**) total amino acid contents during chestnut seed germination. \* represents a significant difference between the treatments and control (CK) at *p* < 0.05. t0: time-point when the seed imbibition was initiated.

#### *3.5. Analysis of the Physiological Response to Germination Following the GABA and VGB Treatments*

We completed a PCA to explore the effects of GABA and VGB before germination (48 h) (Figure 7A) and after germination (120 h) (Figure 7B). Specifically, the physiological traits of the chestnut seeds were evaluated (Table S1). Our data revealed that PC1 and PC2 accounted for 62% and 12% of the physiological variation before germination (48 h), respectively (Figure 7A). The effects of GABA and VGB were clearly separated from the effects of CK by PC1 at 48 h. Additionally, Ser, Pro, Tyr, Val, Met, and Ile were key contributors to PC1, whereas soluble sugars and total amino acids were important factors for PC2 (Table S1). After germination (120 h), PC1 and PC2 accounted for 67% and 14% of the physiological variation, respectively (Figure 7B). The effects of the VGB treatment were clearly separated from the effects of CK; however, the effects of the GABA treatment and CK were uncovered by PC1. These findings indicated that exogenous GABA and VGB were important for inhibiting chestnut seed germination.

**Figure 6.** Effects of exogenous GABA and VGB on free amino acid contents during chestnut seed germination. \* and \*\* represent significant differences between the treatments and control (CK) at *p* < 0.05 and *p* < 0.01, respectively. t0: time-point when the seed imbibition was initiated.

**Figure 7.** Principal component analysis of the effects of exogenous GABA and VGB on physiological parameters at 48 (**A**) and 120 h (**B**).

## **4. Discussion**

Chestnut seeds are widely considered to be healthy for humans [45]. Earlier reports regarding the substantial accumulation of GABA in chestnut seeds [7,36] provided evidence of the health benefits of chestnuts, similar to other GABA-enriched functional foods [9,10]. Additionally, seed germination is closely associated with the shelf-life of chestnuts and with the cultivation of rootstock seedlings [30]. Therefore, it is worth considering the high GABA level regarding its effects on the storage and germination of chestnut seeds.

Recent genetic and physiological studies have implied that GABA is involved in responses to abiotic stresses [46,47], as well as developmental processes, including pollen tube growth [48], primary/adventitious root growth [37,49], and seed germination [20,21]. The recent identification of a GABA receptor, aluminum-activated malate transporter (ALMT) [50], indicates that GABA is a signaling molecule and not just a metabolite [51]. Consequently, GABA functions should be more comprehensively characterized.

Previous studies revealed that endogenous GABA concentrations increase during germination [16,17,52]. In contrast, in the current study, we observed that germination was negatively related to endogenous GABA concentrations at the 48 and 120 h time-points following the GABA and VGB treatments (Figures 1A and 2A). Our results were similar to those of an earlier investigation on high lysine maize seeds [13]. We also detected a more than four-fold increase in GABA concentrations before germination (Figure 2A), which might result from the decrease in Glu content at 48 h (Figure 6) because Glu acts as the direct precursor of GABA production [8,51]. This further indicated that chestnut seeds

may be useful as a GABA-enriched functional food by short-term germination induction. However, the decrease in GABA after 48 h may be ascribed to the requirement of much more Glu for protein synthesis or rapid degradation of GABA back to the TCA cycle during primary root growth [8,51]. However, the application of 10 mM GABA inhibited chestnut seed germination and early primary root growth (Figure 1), which was inconsistent with the results of earlier investigations on barley [20] and *H. ammodendron* [21] seeds. Moreover, blocking GABA degradation with 100 μM VGB also had an inhibitory effect (Figure 1). We recently confirmed VGB to be detrimental to adventitious root growth [37]. Thus, the endogenous GABA may play a specific role in the germination of chestnut seeds and may be useful for improving chestnut seed storage during winter.

Generally, exogenous GABA promotes germination [20,21] and enhances GABA absorption in orthodox seeds [20]. Unexpectedly, our analysis of recalcitrant chestnut seeds uncovered a transient decrease in endogenous GABA concentrations at 48 h following the GABA and VGB treatments (Figure 2A), which was inconsistent with the data generated during our recent investigation of GABA- and VGB-treated poplar stem fragments [37]. Previous studies proved that embryo axes and cotyledons exhibited contrasting responses to desiccation in *Castanea sativa* seeds [27], and the embryonic axes of dormant seeds were maintained in a state of metabolic readiness under optimal conditions [29]. Accordingly, it is possible that the results of the current study were due to the embryos being highly sensitive to the imbibition of exogenous GABA and VGB, resulting in significant increases in endogenous GABA concentrations, relative to the levels of cotyledon with high moisture. However, the whole seed kernels used for measurements may have obscured the final increased GABA concentrations in the embryos. Additionally, GABA provides the carbon skeletons in the TCA cycle [8], which contributes to seed germination [20]. In this study, the GABA and VGB treatments altered the metabolism of soluble sugars, organic acids, and amino acids.

Carbon and nitrogen metabolites, including those mentioned above, are significantly associated with germination and seedling establishment [13], during which GABA critically affects carbon and nitrogen metabolism [8]. During seed germination, as the hormone GA's role [53], exogenous GABA may promote starch hydrolysis to produce soluble sugars by stimulating α-amylase activity [20], but in the current study, there were no significant changes in starch contents following the GABA and VGB treatments (Figure 3B). This may have resulted in the decrease in soluble sugar contents at 120 h (Figure 3A), with the resulting lack of sufficient energy leading to the inhibition of early primary root growth (Figure 1). However, we speculated that the increase in soluble sugar levels at 48 h may have resulted from the lipid breakdown occurring in germinating seeds [53]. We observed that both treatments induced the considerable accumulation of H2O2 (Figure 2B), which can be mainly produced by fatty acid β-oxidation during germination [53], contributing to the inhibition of chestnut seed germination.

The TCA cycle activity is closely associated with seed germination [13], wherein succinate is also the final product of GABA degradation [8]. However, the succinate contents could not be detected in this study. Previous research demonstrated that increasing lysine levels in *Arabidopsis thaliana* seeds resulted in delayed germination, which was accompanied by a significant decrease in the levels of TCA cycle metabolites, such as citrate, malate, and succinate [13]. Hence, the observed decrease in citrate contents at 120 h (Figure 4A) may adversely affect germination and early primary root growth in chestnut seeds because of the associated lack of sufficient energy. Unlike the study by Angelovici et al. [13], we detected an increase in malate contents induced by GABA and VGB treatments (Figure 4B). This increase may block the germination and early primary root establishment of chestnut seeds, which is supported by a recent report [54], which proved that rapidly germinating seeds have low malate levels. Because malate is a key intermediate of the TCA cycle, we speculated that the accumulation of malate may affect the efficient mobilization of storage compounds to supply energy for germination and early seedling development. However, the results of our recent study implied that malate interacting with GABA can delay poplar AR formation [37], possibly because GABA can negatively modulate ALMT via malate [50,55,56]. Thus, we considered that changes to the metabolic

status of malate and GABA led to physiological responses, such as the delayed seed germination and inhibited early primary root growth observed in this study, through modulated ALMT activities.

In many plant species, most amino acids accumulate during seed germination [12]. For example, the aspartic acid family of amino acids contributes to the onset of autotrophic growth-associated processes during germination [13]. An exogenous nitric oxide donor (*S*-nitroso-*<sup>N</sup>*-acetyl-<sup>d</sup>,<sup>l</sup>-penicillamine) can enhance the germination of Kabuli chickpea seeds, which coincides with an increase in amino acid levels [57]. Exogenous H2O2 also promotes the germination of eggplant seeds, accompanied by enhanced amino acid biosynthesis and protein expression [58]. Therefore, the observed decrease in most of the amino acid levels at the 48 h time-point following the GABA and VGB treatments might contribute to the inhibition of chestnut seed germination (Figure 6). At 120 h, however, only the VGB treatment induced a considerable increase in the accumulation of 10 amino acids (Figure 6), which were reportedly negatively associated with root growth [13,37,58–60]. Thus, VGB might have a specific role in inhibiting early primary root growth. However, the GABA treatment did not adversely a ffect the roots, which was consistent with the findings of a previous study involving *Brassica napus* seedlings [60], but it inhibited chestnut seed germination. Furthermore, soluble protein contents reportedly decreased during the seed germination of six grass species [61], but they increased significantly when the primary root growth was inhibited at 120 h after both treatments in the current study (Figure 5). Therefore, the two treatments appeared to negatively influence early primary root growth in chestnut seeds.
