Rapid intracellular acidification is a novel plant defense response countered by the brown 1 planthopper 2

of rice suggests that NICA is a promising target for


INTRODUCTION
The brown planthopper (BPH; Nilaparvata lugens Stål, Hemiptera, Delphacidae) is a monophagous insect pest of rice (Oryza sativa L.) found in all rice-growing Asian countries.The BPH sucks rice phloem sap via its stylet, causing leaf yellowing and wilting, stunted plant growth, reduced photosynthesis and ultimately death of rice plant 1 .During severe BPH outbreaks, tens of thousands of insects swarm on a rice field, resulting in the 'hopperburn' phenomenon, which is characterized by large-scale wilting, yellowing and lethal drying of rice plants 2 .Besides direct damages, the BPH may also indirectly damage rice plants by oviposition and transmitting viral disease agents 3,4 .Application of chemical insecticides has been a main strategy for controlling BPH.Although it has the advantages of having rapid effects on killing insects and low costs, use of insecticides leads to environmental pollution and resistance of BPH to pesticides.In recent decades, breeding rice resistant varieties to control BPH has attracted increasing attention.To date, more than 30 resistance genes have been found in the rice genome 5 .However, BPH often quickly evolves new biological types that evade rice resistant genes.Therefore, additional BPH controlling methods need to be developed to complement the current control measures toward long-term solutions of achieving durable BPH resistance in rice.One method could be based on disruption of key steps of BPH's natural infestation process.However, development of such methods will require a comprehensive understanding of the basic biology of the BPH-rice interaction.
A critical step in BPH infestation of rice is secreting bioactive substances into the plant tissues through the stylet 6-8 . Specifically, during feeding, BPH secretes both colloidal and watery saliva 9,10 .The main function of the colloidal saliva is to form a saliva sheath around the piercingsucking mouthparts, stabilizing the overall feeding apparatus.The composition and function of watery saliva is more complex, and it contains salivary proteins that are believed, in most cases, to regulate various pathways in plant cells to enhance BPH feeding and survival in rice plants.For example, NlEG1, a salivary endo-b-1,4-glucanase, degrades plant celluloses to help the BPH's stylet reach to the phloem 11 .NlSEF1, an EF-hand Ca 2+ -binding protein, interferes with calcium signaling and H 2 O 2 production during BPH feeding 12 .Salivary protein 7 is required for normal feeding behavior and for countering accumulation of a defense compound, tricin 13 .On the other hand, mucin-like protein (NlMLP) triggers defense responses in rice cells, including cell death, callose deposition and up-regulation of pathogen-responsive genes 14,15 .
Carbonic anhydrases (CAs) (EC 4.2.1.1)are zinc metalloenzymes that function as catalysts in the bidirectional conversion of CO 2 and water into bicarbonate and protons 16   .There are at least five distinct CA families (α-, β-, γ-, δ-, and ε-CAs) and three of them (α-, β-, and γ-CAs) are ubiquitously distributed among animal, plant, and bacterial species.The widespread distribution and adequate abundance of these CA families underline their evolutionary importance throughout the kingdoms of life.CAs participate in a wide range of biological processes, such as pH regulation, CO 2 homeostasis, stomatal aperture, and plant defense 17-21   .NICA belongs to the α-CA subfamily.Our previous study showed that NlCA expressed in BPH salivary glands 6 .Surprisingly, however, RNA interference (RNAi) of the NlCA transcript in BPH insects affects neither pH maintenance within the salivary gland, watery saliva or gut, nor insect feeding behavior or honeydew excretion, but greatly reduced survival of BPH on rice plants, suggesting a critical function in planta via an unknown mechanism 6 .
Here, we report that NlCA-RNAi BPH feeding results in rapid intracellular acidification of rice cells.We found that NlCA is secreted into the rice tissues and functions as an effector that stabilizes host cell intracellular pH, accompanied by suppression of defense responses, during BPH feeding.Thus, we have uncovered intracellular pH homeostasis is a previously uncharacterized battleground in plant-insect interactions.

Results
NlCA is detected in rice sheath tissues during BPH feeding.NlCA was previously found to be highly expressed in salivary glands and present in the watery saliva of BPH fed on artificial diet 6 .
We conducted a more detailed characterization of NICA expression in this study.RNA in situ hybridization showed that the expression level of NlCA was detectable throughout the principal glands (PGs) and accessory glands (AGs), but not in A-follicle of the principal gland (APG) (Figure 1A), which further raised the possibility that NICA may be one of the "effector proteins" secreted into the rice tissue during BPH feeding on rice plants.To test this possibility, we compared protein profiles in the leaf sheaths of Nipponbare rice plants before and after BPH feeding using liquid chromatograph-mass spectrometer (LC-MS).We found 8 NICA-specific peptides in BPH-fed leaf sheath tissue (Figure 1B), confirming that NlCA is secreted into the host tissues during BPH feeding.As BPH is a phloem-feeding insect, its effector proteins, such as NlCA, presumably act in the phloem.We therefore collected the phloem exudate from Nipponbare rice leaf sheaths with or without BPH feeding for LC-MS analysis (Figure 1C).Six NlCA-specific peptides were detected in the phloem exudate of Nipponbare plants fed by BPH (Figure 1B), demonstrating that NlCA is delivered into phloem by BPH.

Transgenic expression of
NlCA in rice rescues the ability of NlCA-silenced BPH to feed and survive.To further clarify the site of function (i.e., in insect vs. in plant) of NlCA in the BPHrice interaction, we produced transgenic Nipponbare plants expressing NICA (see Methods).A total of 26 lines were produced and 6 lines were found to robustly express the NICA transcript (Supplemental Figure 1A).NlCA-expressing plants exhibited no noticeable changes in appearance compared to Nipponbare plants (Supplemental Figure 1B-1D).NlCA-expressing plant lines were propagated to T3 generation, and three lines were subjected to further characterization, including BPH feeding.For BPH feeding assay, double-stranded RNA (dsRNA) of NlCA (dsNlCA) or the control green fluorescent protein gene (dsGFP) was injected into 3 rd instar BPH nymphs to initiate RNAi of the NICA transcript 22 .Quantitative real-time PCR analysis confirmed that the transcript levels of the NlCA gene were reduced by 99% and 97%, respectively, in tested individuals when compared with non-RNAi control and dsGFP-treated insects (Figure 2A).There were no significant differences in the survival rates between dsGFP and dsNlCA BPH when fed on the artificial diet, indicating that silencing NlCA expression has no obvious impact on the basic physiology of BPH (Figure 2B).However, we found that the survival rate of dsNlCA BPH was sharply decreased, starting at day 9 post-infestation, to ~40% at day 14, whereas the control dsGFP BPH survived normally on wild-type Nipponbare plants (Figure 2C).This result is consistent with previous results conducted in the japonica cultivar Xiushui134 rice plants 6 , suggesting that the requirement of NICA for BPH survival is not specific to a specific rice genotype.Strikingly, NICA transgenic plants almost fully restored the survival of NlCA-silenced BPH insects (Figure 2C; Supplemental Figure 1E-1F), demonstrating that NlCA expressed in the host tissue (Figure 2D and 2E) can complement the infestation defect of NlCA-silenced BPH insects.
The conserved catalytic site amino acids of NlCA are critical to its function in rice.NlCA contains several conserved amino acid residues predicted to be at the catalytic site of carbonic anhydrases (Figure 2F).We asked if some of these conserved active site residues are critical to the function of NlCA in the rice-BPH interaction.Accordingly, we expressed three different NlCA mutants in Nipponbare plants (Supplemental Figure 1A   Interestingly, we noted that the control dsGFP wild-type BPH feeding induced slight intracellular acidification at 12 h (Figure 3E).At this late time point of feeding, most BPHs would have left their initial feeding sites and have relocated to new feeding sites.In our analysis, we could not distinguish between old and new feeding sites.It is therefore likely that active BPH feeding (i.e., active NlCA injection) is needed to maintain intracellular pH homeostasis at all feeding sites.
Next, we conducted experiments to determine if intracellular acidification is a highly localized celltype specific response or a spreading local response.We imaged and calculated the intracellular pH in the mesophyll cells and epidermal cells at the feeding site.We found that intracellular acidification occurred in the mesophyll cells and epidermal cells (Supplemental Figure 2I-2L) in response to feeding by dsNlCA BPH, albeit to a lesser degree compared to that in the sieve elements/companion cells.In contrast, rice plants fed by dsGFP wild-type BPH responded with an initial slight intracellular acidification at 1.5 h, but then consistently maintained a more alkalized intracellular pH in the mesophyll and epidermal cells compared to non-fed rice plants and dsNICA BPH (Supplemental Figure 2I-2L).
Because extracellular pH change is associated with plant responses to biotic stress  3A-H).The pHusion sensor plants exposed to external pH of 2 showed intracellular acidification that simulates pH changes observed during BPH feeding (Figure 3J-K).Survival rates of BPH insects were lower in Nipponbare plants growing in media with pH of 2, compared to those growing in media with pH of 4 (Figure 3L).Rice growth medium of pH 2 was used to attain an intracellular pH similar to that during BPH feeding (Figure 3K).These results suggest that cellular acidification is sufficient to induce defense gene expression in rice and to reduce BPH's ability to survive on rice.
Rice defense responses are inhibited by NlCA.As NICA stabilizes intracellular pH during WT BPH feeding (Figure 3A-D) and both dsNlCA BPH feeding and ectopic intracellular acidification causes activation of defense gene expression (Figure 3F-I; Supplemental Figure 3), we next tested the hypothesis that a major function of NICA-mediated stabilization of intracellular pH may be to prevent over-stimulation of downstream rice defense responses during feeding.Callose deposition in the phloem sieve tubes is a classical defense response that is associated with feeding of piercing-sucking insects

35
. We examined this response using aniline blue to stain callose in the phloem sieve cells.Indeed, we found that fewer and smaller callose deposition was found in the sieve plates of NlCA-expressing leaf sheaths compared to those found in the sieve plates of Nipponbare fed by BPH (Figure 4A-H) at 72 h after BPH feeding, demonstrating that NlCA suppresses callose deposition in sieve cell pates.Furthermore, dsNlCA BPH induced higher expression of callose biosynthesis genes, such as OsGSL1, OsGSL3, OsGLS5 and OsGns5, than the dsGFP control BPH; however, the induction of these genes was greatly compromised in NlCA transgenic plants with or without BPH treatment (Figure 4I-L).We also measured the transcript levels of defense marker genes (e.g., OsNH1, OsNH2, OsWRKY45 and Collectively, these results showed that NICA-mediated intracellular pH homeostasis is linked to downregulation of callose deposition in phloem sieve cells as well as defense gene expression in rice plants.

Discussion
In this study, we provided evidence that intracellular acidification is a previously unrecognized plant defense response that occurs during BPH feeding on rice.This finding was facilitated by our attempt to understand the role of NICA in the rice-BPH interaction.We found that the NlCA transcript is detected mainly in the salivary glands (Figure 1A) and that the NICA protein is found in rice tissues, including the phloem sap, fed by BPH (Figure 1B).NlCA-silenced (dsNlCA) insect survived very poorly on at least two independent cultivars of rice plants, Xiushui134 6 and Nipponbare (Figure 2), whereas NlCA expressing rice plants can restore the normal survival of NlCA-silenced (dsNlCA) insects (Figure 2; Supplemental Figure 1), suggesting that NlCA functions in plant cells.Using the cytoplasm pH sensor, we found that NlCA is required for BPH to maintain a normal plant cytoplasm pH during BPH feeding (Figure 3A-D; Supplemental Figure 2).
Pathogen/insect-derived effectors can be powerful molecular probes to discovering novel plant regulators/responses to biotic attacks.Although other piercing/sucking herbivores derived effectors, such as Btfer1, LsPDI1, LsSP1, Mp55, DNase II and BISP, have been reported as defense-suppressive effectors [36][37][38][39][40][41] , our discovery of intracellular acidification as a plant defense response and our finding that brown planthopper (BPH) secretes NICA as a counter-defense measure illustrates host cell pH homeostasis as a novel battlefield that has not been revealed in any plant-biotic interactions.
Extracellular alkalinization of culture plant cells has long been recognized as a canonical plant response to microbial elicitors as well as endogenous plant signals Because cellular pH alterations could potentially affect multiple biomolecules and, hence, multiple cellular processes, future research should comprehensively define all cellular processes that are affected by intracellular pH acidification.In this study, we found that this pH change is linked to activation of callose deposition at phloem sieve cells and expression of defense response genes, such as OsNH1, OsNH2, OsPBZ1 and OsWRKY45 (Figure 3; Supplemental Figure 3).In reverse, NICA-mediated intracellular pH stabilization dampens these defense responses (Figure 4).Callose deposition in phloem sieve cells, in particular, is a classical defense response to a variety of sucking/piecing insects and is thought to limit nutrient flow during insect feeding 35,[45][46][47][48] .
Together, these results suggest that a major effect of NlCA-mediated pH stabilization is to prevent overactivation of defense responses during BPH feeding.
The mechanism by which NICA counters intracellular acidification is likely inherent to its reversible inter-conversion of carbon dioxide and water into carbonic acid, protons and bicarbonate ions.CAs are universally present in all organisms (Supplemental Figure 4); other piecing/sucking insects may use CAs or another mechanism to manipulate host intracellular pH as part of their infestation strategy.Indeed, CA has been reported as a protein component of saliva in rice green leafhopper, Nephotettix cincticeps 49 , and aphid Myzus persicae 50 .In the case of M. persicae, CA-II was shown to increase viral transmission via plant apoplastic acidificationmediated acceleration of intracellular vesicular trafficking 50 .However, because NICA plays a critical role in BPH's survival on rice plants per se (i.e., in the absence of viral infection), as shown in this study, it is more likely that insect-secreted CAs constitute play a primary role in facilitating insect survival by countering intracellular acidification-associated defense activation.In fact, future research should examine if CA-mediated increase in viral transmission may be a secondary consequence of defense suppression, which was not examined in the previous study 50 .
Discovering the role of pH regulation during plant response to biotic and abiotic stresses and characterizing the impact of such pH alterations could be an important area for future research.
Because maintaining proper external and internal pH is critical for all forms of life, prokaryotic and eukaryotic organisms, alike, have evolved mechanisms to achieve pH homeostasis.Facing the fluctuating external pH, prokaryotes have evolved diverse mechanisms for sensing external pH.
For instance, the bimodal sensing of pH is employed by Bacillus subtilis and Escherichia coli

Competing Interest Statement:
The authors declare no competing interest.benthamiana leaves and rice protoplasts were as negative controls (N).Ponceau S staining of Rubisco shows protein loading control.See also Table S1.Different letters indicate statistically significant differences analyzed by two-way ANOVA (Tukey test, P < 0.05).Experiments were repeated three times with similar trends.

STAR METHODS
Detailed methods are provided in the online version of this paper and include the following: • KEY RESOURCES

Figure 1B - 23 .
Figure 1B-1D).We conducted BPH survival assay and found that, in contrast to plants expressing wild-type NICA, the NlCA-m2 and NlCA-m3 plants could not fully rescue the infestation defect of NlCA-silenced BPH insects.The NlCA-m1 plants, on the other hand, could recover the infestation defect of dsNICA BPH (Figure2Gand 2H; Supplemental Figure1H).Thus, the conservative residues Glu182 and His184 at the predicted catalytic site are indispensable for the function of NlCA inside the plant cell.

OsWRKY13)
in Nipponbare and NlCA-expressing plants fed by dsGFP and dsNlCA BPH.As shown in Figure 4M-P, induced expression of OsNH1, OsNH2, OsWRKY13 and OsWRKY45 by dsNlCA BPH was suppressed in NlCA transgenic plants compared to those in Nipponbare plants.

26 . 44 .
Extracellular alkalinization caused by plant endogenous RAPID ALKALINIZATION FACTOR (RALF) peptides, for example, are perceived by the FERONIA-family receptors42,43 .There is evidence that this perception causes phosphorylation of PM-localized H(+)-ATPase 2, resulting in the inhibition of proton transport across the PM Notably, extracellular alkalinization has recently been shown to inhibit or promote growth-or immunity-associated cell surface receptor functions through specific pHsensitive amino acid sensors26 .In contrast, defense regulation by intracellular pH changes has so far escaped the discovery by researchers until this study.Our demonstration of a link between intracellular acidification and defense activation has laid a foundation for future discovery of potentially diverse pH-sensitive intracellular regulators of defense responses, which could add a new dimension in the study of plant-biotic interactions.

51 .
Fungi employ a conserved pathway, mediated by Rim101 and PacC, to sense external pH52 .It has been reported that different subcellular compartments within the plant cell maintain different pH values, presumably as part of carrying out their unique physiological functions 53 .The demonstrated ability of NICA to counter stimulus-dependent pH changes in plant cells could make NICA a useful molecular tool to modulate and broadly understand the effects of pH stabilization on plant signaling and metabolic pathways in different cell types, organelles, and tissues in plants by, for example, targeting NICA expression in specific tissues, cells, or organelles.Additionally, the crucial role of NICA for BPH infestation of rice suggests that NICA is an important target for chemical or trans-kingdom RNAi-based inactivation for the development of novel BPH control strategies in plants.experiments and analyzed data; C.X.Z. and S.Y.H. analyzed data.Y.J.J., X.Y.Z., C.X.Z. and S.Y.H. wrote the paper with all authors approved the final article.

Figure 1 .
Figure 1.Initial characterization of Nilaparvata lugens carbonic anhydrase (NlCA).(A) in situ RNA hybridisation of salivary glands in 5-instar BPHs.Left, a schematic diagram of BPH salivary glands, including the principal glands (PG), the accessory glands (AG) and the Afollicle of the principle gland (APG).Right, NlCA expression was detectable in the PG (red) using antisense NICA sequence as a probe.Sense NICA probe was used as a negative control.Nuclei are stained blue by 4',6-diamidino-2-phenylindole (DAPI).Scale bar = 50 μm.(B) The amino acid sequence of NlCA.The highlighted amino acid residues indicate the peptides detected in BPHinfested rice sheath tissue by LC-MS analysis.The underlined amino acid residues indicate the peptides detected in phloem exudate of BPH-infested rice by LC-MS analysis.(C) A schematic diagram of the phloem exudate collection.(D) NlCA is colocalized with the YFP signals in N. benthamiana (upper row; Scale bar = 10 μm) and rice cells (lower row; Scale bar = 5 μm).YFP and NlCA-CFP fusion proteins were co-expressed in N. benthamiana leaf cells for 48 h using the Agrobacterium-mediated transient expression method.YFP and NlCA-CFP fusion proteins were co-expressed in rice protoplasts 16 h after the corresponding DNA constructs were introduced into rice protoplasts via polyethylene glycol-mediated transformation.Experiments were repeated three times with similar trends.(E) NlCA-CFP-HA fusion protein levels are detected with anti-HA (Zenbio, 301113) in N. benthamiana leaves and rice protoplasts.Protein samples were extracted from NlCA-CFP-HA expressing N. benthamiana and rice protoplasts.Proteins from mock N.

Figure 2 .
Figure 2. Effects of NlCA RNAi on BPH survival on rice cultivar Nipponbare.(A) The NICA transcript levels in BPH at 3 dpi post-injection of dsGFP or dsNICA were determined by qRT-PCR (displayed as % of the NICA transcript abundance in control BPH).Values are displayed as mean ± SEM of 4 experimental replicates (Two-way ANOVA; 3 biological replicates in each experiment and 30 individual insects pooled for each biological replicate).(B) The daily survival rates of dsNlCA-injected BPH insects feeding on artificial diet.Values are displayed as mean ± SEM of 3 biological replicates (one biological replicate includes 50 individual BPH adults fed on artificial diet in a lucifugal plastic bottle).(C) The daily survival rates of dsNlCAinjected BPH insects feeding on Nipponbare and NlCA-expressing line 1.Values are displayed as

Figure 3 .
Figure 3. Role of NlCA in maintaining intracellular pH of rice cells.(A) Confocal microscopic images at BPH feeding sites in Nipponbare plants expressing a ratiometric cytoplasmic pH sensor (Cyto-pHusion) 4 h after placing 30 5 th -instar BPH nymphs on each plant.A 40.0 x objective was used to capture every feeding site on the tiled 1 cm x 1 cm section of rice leaf sheath.Confocal images from plants with no BPH feeding served as the control.Scale bar = 20 μm.(B-E) EGFP:mRFP signal ratios at 1.5 h (B), 4 h (C), 8 h (D) and 12 h (E) of dsGFP or dsNlCA BPH treatment compared with no BPH control.EGFP was imaged at λ Ex = 500 nm and λ Em = 540 nm.mRFP was imaged at λ Ex = 570 nm and λ Em = 620 nm.Values are displayed as mean ± SEM (n ≥ 24 circular areas of leaf sheath phloem with a diameter of 100 μm with the feeding site at the center).(F-I) Expression of defense response genes, OsWRKY45 (F), OsWRKY13 (G), OsNH1 (H) and OsNH2 (I), in Nipponbare plants infested by dsGFP or dsNlCA BPH.Values are displayed as mean ± SEM of three biological replicates.Each biological replicate represents pooled leaf sheaths from three individual rice plants fed by 20 5 th instar BPH nymphs per plant.(J) Confocal microscopic images of Nipponbare plants expressing a ratiometric cytoplasmic pH sensor (Cyto-pHusion) at 12 h after being transferred to Yoshida medium at pH of 2. Confocal images from plants grown in Yoshida medium with pH of 4 served as the control.Scale bar = 15 μm.(K) EGFP:mRFP signal ratios.EGFP was imaged at λ Ex = 500 nm and λ Em = 540 nm.mRFP was imaged at λ Ex = 570 nm and λ Em = 620 nm.Values are displayed as mean ± SEM (n ≥ 33 calculation area per condition).(L) The 7 th -day survival rates of dsGFP-and

Figure 4 .
Figure 4. NICA dampens callose deposition and defense gene expression.(A-F) Callose accumulation (bright blue fluorescence indicated by red arrows) on the sieve plates of Nipponbare leaf sheaths (A and D) and NlCA-OE leaf sheaths (B and E) plants at 72 h after BPH feeding.P, phloem.The pictures were taken under a Zeiss Microscope.(A-B) Crosssections.Scale bar = 5 μm.(D-E) Longitudinal sections.Scale bar = 10 μm.(C and F) The bright field views of cross and longitudinal phloem sections, respectively.(G) Total areas of callose deposition in BPH-infested leaf sheaths of Nipponbare and NlCA-OE.Each data point represents the total areas of callose deposition found in 300 cross-sections of each experiment.Values are displayed as mean ± SEM of four experiments.(H) Total number of callose deposits in BPHinfested leaf sheaths of Nipponbare and NlCA-OE.Each data point represents the total number of callose deposits found in 300 cross-sections of each experiment.Values are displayed as mean ± SEM of four experiments.(I-L) Relative expression levels of the callose synthase gene OsGSL1 (I), OsGSL3 (J), OsGSL5 (K) and OsGns5 (L) in response to BPH feeding.Values are displayed as mean ± SEM of 3 biological replicates.Each biological replicate represents pooled leaf sheath form three individual rice plants fed by 20 5 th instar BPH nymphs per plant.(M-P) Expression of defense marker gene OsWRKY45 (M), OsWRKY13 (N), OsNH1 (O) and OsNH1 (P) in response to BPH feeding.Values are displayed as mean ± SEM of 3 biological replicates.Each biological replicate represents pooled leaf sheaths from three individual rice plants fed by 20 5 th instar BPH nymphs per plant.ns indicates no significant difference between treatments (Two-way ANOVA).

○F
Callose deposition in rice sheath• QUANTIFICATION AND STATISTICAL ANALYSISI H C H W G R N D I M G S E H F I D A E S Y P M E C H M ----I A C A W G R N D I M G S E H F I D A E S Y P M E C H M ----I H C H W G R N D I M G S E H F I D A E S Y P M A C A M ----I A C A W G R N D I M G S A H F I D A E S Y P M A C A M ----158 185 L T T P P F T E A V T F I L T T P P F T E A V T F I L T T P P F T E A V T F I L A T P P F T E A V TA I RGGPMAPSEP EFPTELEGED RFLVNPLNVD EVLSFGPRAP QGSSPGFPWD RSQLPQPQQS PIDICRCKCK SMTYQPLKLS YAKTSRNDGI VVEISNSGHG VNLKVPKAEA KPFLSGGMLS NRYIYDNIHC HWGRNDIMGS EHFIDAESYP MECHMVTFNE KYGTINEALK NKQDGVVVLT WFYKLANEDN KRLQNFLQYV PSVRENGAKI FRADQNALFF PLEVLPINQE YFAYPGSLTT PPFTEAVTFI ILPIAVGISK DQLALLRTLM DDKNPSKPVL SNRREIQTLN HRTVYCVKGR VQAISREIPP QQSNDVYNQI HRYC