Nodulin 26-like intrinsic protein CsNlP2;2 is a silicon influx transporter in Cucumis sativus L.

DUAN Yao-ke,SU Yan,HAN Rong,SUN Hao,,GONG Hai-jun

1 Shaanxi Engineering Research Center for Vegetables/College of Horticulture,Northwest A&F University,Yangling 712100,P.R.China

2 Henan Key Laboratory of Ion-Beam Bioengineering,School of Agricultural Sciences,Zhengzhou University,Zhengzhou 450000,P.R.China

Abstract Nodulin 26-like intrinsic proteins (NIPs) are a family of channel-forming transmembrane proteins that function in the transport of water and other small molecules. Some NIPs can mediate silicon transport across plasma membranes and lead to silicon accumulation in plants,which is beneficial for the growth and development of plants. Cucumber is one of the most widely consumed vegetables;however,the functions of NIPs in this crop are still largely unknown. Here,we report the functional characteristics of CsNIP2;2. It was found that CsNIP2;2 is a tandem repeat of CsNIP2;1,which had been demonstrated to be a silicon influx transporter gene. CsNIP2;2 has a selectivity filter composed of cysteine,serine,glycine and arginine (CSGR),which is different from all previously characterized silicon influx transporters in higher plants at the second helix position. Xenopus laevis oocytes injected with CsNIP2;2 cRNA demonstrated a higher uptake of silicon than the control,and the uptake remained unchanged under low temperature. CsNIP2;2 was found to be expressed in the root,stem,lamina and petiole,and exogenous silicon treatment decreased its expression in the stem but not in other tissues. Transient expression of CsNIP2;2-eGFP fusion sequence in onion epidermal cells showed that CsNIP2;2 was localized to the cell nucleus,plasma membrane and an unknown structure inside the cell. The results suggest that CsNIP2;2 is a silicon influx transporter in cucumber,and its subcellular localization and the selectivity filter are different from those of the previously characterized silicon influx transporters in other plants. These findings may be helpful for understanding the functions of NIPs in cucumber plants.

Keywords:cucumber (Cucumis sativus L.),nodulin 26-like intrinsic membrane protein (NIP),silicon influx transporter,aromatic/arginine selectivity filter

1.Introduction

Aquaporins are membrane proteins that are members of the major intrinsic proteins (MIPs),and they are involved in the transport of water and other small molecules (Maurelet al.2008). According to their localization and sequence homology,plant aquaporins can be divided into seven subfamilies,including plasma membrane intrinsic proteins(PIPs),tonoplast intrinsic proteins (TIPs),nodulin-26-like intrinsic proteins (NIPs),small basic intrinsic proteins(SIPs),hybrid intrinsic proteins (HIPs),X intrinsic proteins(XIPs),and GlpF-like intrinsic proteins (GIPs) (Maurelet al.2015). Among these aquaporin subfamilies,NIPs play an irreplaceable role in plants because they can facilitate the transport of water,silicon,boron and other small molecules across membranes (Mitani-Uenoet al.2011;Pommerreniget al.2015). According to the sequence similarity of the aromatic/arginine (ar/R)region,NIPs can be categorized into three groups -NIP I,NIP II and NIP III -with different transport substrates(Wallace and Roberts 2004). NIP I can transport water,glycerol and lactic acid,while NIP II can transport boron and urea;whereas NIP III proteins have transport activities for silicon,arsenic and selenium (Deanet al.1999;Wallace and Roberys 2005;Maet al.2006,2008;Takanoet al.2006;Choi and Robert 2007;Zhao F Jet al.2010;Zhao X Qet al.2010;Mitani-Uenoet al.2011).

Similar to other aquaporin proteins,NIPs have six putative membrane-spanning helix domains and five loops(named A–E from the N terminal of aquaporin) (Maurelet al.2008). There is a highly conserved Asn-Pro-Ala(NPA) motif on loop B and loop E,respectively;and the two motifs form a narrow pore constriction for selective transport (Maurelet al.2008). Besides the NPA region,NIPs also have an ar/R selectivity filter,which is formed by four residues from helix 2 (H2),helix 5 (H5) and loop E (LE1 and LE2) (Maurelet al.2008;Pommerreniget al.2020).

The functions of NIPs have been investigated in some species,especially model plants. For instance,AtNIP1;1 inArabidopsisthalianaexhibits transport activities for trivalent arsenic and antimony (Kamiyaet al.2009;Kamiya and Fujiwara 2009). AtNIP1;2 is permeable to aluminum-malate and trivalent arsenic(Bienertet al.2008;Wanget al.2017). AtNIP3;1 can enhance trivalent arsenic uptake,and AtNIP5;1,AtNIP6;1 and AtNIP7;1 have transport activities for both boron and trivalent arsenic (Wallace and Robert 2005;Bienertet al.2008;Isayenkov and Maathuis 2008;Li Tet al.2011;Gomezsotoet al.2019). In rice,OsNIP1;1 is permeable to trivalent arsenic (Bienertet al.2008). OsNIP2;1 can mediate the uptake of arsenite,methylated arsenite,selenite and silicic acid (Maet al.2008;Liet al.2009;Zhao X Qet al.2010). OsNIP2;2 has a transport activity for silicon,and OsNIP3;1 facilitates the transport of boric acid (Yamajiet al.2008;Yamaji and Ma 2009;Hanaokaet al.2014).

Cucumber (CucumissativusL.) is one of the most widely consumed vegetable crops. There are nineNIPgenes in the cucumber genome (Zhuet al.2019). While CsNIP2;1 shows transport activities for both urea and silicon (Zhanget al.2016;Sunet al.2017),the transport activities and functions of other CsNIPs in cucumber still remain unknown. In this study,we analyzed theNIPgenes in cucumber and isolatedCsNIP2;2,which was proven to be a silicon transporter gene,and CsNIP2;2 has a different subcellular localization and selectivity filter from those of the previously characterized silicon influx transporters in other plants. The study may help us to better understand the functions of NIPs in cucumber plants.

2.Materials and methods

2.1.Plant materials

Cucumber ‘Mch-4’ seeds were obtained from the Cucumber Research Team at Northwest A&F University,China. The seeds were sterilized and seedlings were cultured as described in our previous publication (Sunet al.2017). Briefly,the cucumber seeds were germinated in the dark at 28°C for 2 d and then cultured in sterilized substrate at a temperature cycle of 26°C/18°C,14 h/10 h. Seedlings at the 3rd-leaf age were transplanted into plastic boxes containing diluted Hoagland solution(Hoagland and Arnon 1950). The culture solution contained 1.25 mmol L–1KNO3,1.25 mmol L–1Ca(NO3)2,0.5 mmol L–1MgSO4,0.25 mmol L–1KH2PO4,46.1 μmol L–1H3BO3,0.24 μmol L–1(NH4)6Mo7O24,9.1 μmol L–1MnCl2,0.76 μmol L–1ZnSO4,0.32 μmol L–1CuSO4and 71 μmol L–1FeNa-EDTA. The pH of the solution was adjusted to 6.0 daily,and the solution was renewed every 5 d.

2.2.Bioinformatic analysis of NIPs

The physical location information of cucumberNIPswas obtained from the cucumber genomic database with TBtools (Chenet al.2020),and allNIPgenes were mapped to the chromosomes. The deduced amino acid sequences were aligned using DNAMAN v.6 (Lynnon Biosoft,San Ramon,CA,USA),and the phylogenetic tree was constructed using the Maximum Likelihood method with MEGA X (Kumaret al.2018). The tree with the highest log likelihood is shown below. Initial tree(s)for the heuristic search were obtained automatically by applying the Neighbor-Join and BioNJ algorithms to a matrix of estimated pairwise distances. The tree is drawn to scale,with branch lengths measured by the number of substitutions per site. All positions with less than 95% site coverage were eliminated,i.e.,those with fewer than 5%alignment gaps,missing data,and ambiguous bases were allowed at any position (partial deletion option).

2.3.Isolation of CsNIP2;2

The cucumber roots were sampled and frozen in liquid nitrogen 7 d after being transplanted. Total RNA was extracted from the roots using a plant RNA kit (Omega,Madison,WI,USA). The first strand cDNA was synthesized with a HiScript III 1st Strand cDNA Synthesis Kit (Vazyme,Nanjing,China) according to the instructions provided by the manufacturer. The coding sequence(CDS) ofCsNIP2;2was amplified using PrimerStar max polymerase (TaKaRa,Kusatsu,Japan) with a pair of primers (forward:ATGGCGAGACGCAGCGACGATGA;reverse:TCAATGTAACGAGAGTGAATGGGTG) based on the cucumber genomic information (Huanget al.2009;Li Zet al.2011).

2.4.Transport activity determination

To determine the transport activity of CsNIP2;2 for silicon,we expressedCsNIP2;2inXenopuslaevisoocytes.The CDS ofCsNIP2;2was amplified with primers (F:GAAGATCTATGGCGAGACGCAGCGACGATGA;R:GAAGATCTTCAATGTAACGAGAGTGAATGGGTG) and then inserted into theBglII site of the pXBG-ev1 vector(kindly provided by Dr.Maki Katsuhara from Okayama University,Japan). The construct was linearized by restriction enzyme digestion at theXbaI restriction site and complementary RNA (cRNA) with a cap analogue was synthesized with a Riboprobe?System-T3 and Ribo m7G Cap Analogue (Promega,Madison,WI). Oocyte defolliculation,culture conditions and selection were as described previously (Mitaniet al.2008). The determination of the transport activity was conducted according to Sunet al.(2020) using germanium -a silicon analogue (Nikolicet al.2007) -as the substrate. Briefly,a total of 50 nL cRNA (1 ng nL?1) was injected into the oocytes using an MSM System (Nikon,NT-88-V3,Japan). As a negative control,RNA-free water was injected instead. After incubation in modified Birth’s saline solution (MBS,88 mmol L–1NaCl,1 mmol L–1KCl,2.4 mmol L–1NaHCO3,15 mmol L–1HEPES at pH 7.6,0.33 mmol L–1Ca(NO3)2,0.41 mmol L–1CaCl2,0.82 mmol L–1MgSO4,10 μg mL?1sodium penicillin,and 10 μg mL?1streptomycin sulphate) for 24 h at 18°C,the oocytes were incubated in MBS containing 1 mmol L–1germanium oxide for 30 min.Then the oocytes were carefully washed with MBS and the amount of germanium (Ge) in the oocytes was determined by ICP-MS (ICAP-Qc Thermo Fisher Scientific,Waltham,MA).

2.5.Promoter characteristics and expression pattern analysis

Cis-acting elements in the promoter region ofCsNIP2;2(1 500 bp) were analyzed using PlantCARE (Lescotet al.2002).

The qPCR technique was used to determine the expression patterns ofCsNIP2;2in different tissues of cucumber seedlings. Six hours after being transferred to hydroponic culture with or without added silicon (1 mmol L–1silicic acid),the new fully expanded leaves,stems and roots were collected and frozen in liquid nitrogen for expression analysis. Total RNA and first strand cDNA were synthesized as described above. Relative qPCR was performed with a SYBR Premix EX TaqTM(TaKaRa,Kusatsu,Japan) using an ABI StepOne plus qPCR System (Applied Biosystems,Carlsbad,CA). The specific primers forCsNIP2;2were as follows:forward,GAACCCAGCAAGATCAATAGGAC;and reverse,TAACTGGACCGATCATGTACACC. Ubiquitin extension protein (UBI-ep) was introduced as an internal control with a forward primer of CCAAAGCACAAGCAAGAGAC,and a reverse primer of AGTAGGTTGTCTTATGGCGC (Wanet al.2010). Six biological replicates were used for these analyses (n=6).

2.6.Intracellular localization of CsNIP2;2

The CDS ofCsNIP2;2was amplified with primers (forward:ACGCGTCGACATGGCGAGACGCAGCGACGATGA;reverse:ACGCGTCGACATGTAACGAGAGTGAATGGGTGTG)and inserted into the pGEM-T Easy Vector (Promega,Madison,WI,USA). After sequencing by Sangon Biotech Co.,Ltd.(Shanghai,China),the CDS was cut from the pGEM-T Easy Vector and inserted into pTF486 atSalI andBamHI sites with the 3′ ends fused with enhanced green fluorescent protein (eGFP) sequences. Expression ofCsNIP2;2in onion epidermal cells was performed by gene gun transduction as described by Scottet al.(1999). Onion epidermal cells transformed with empty vector (pTF486) was introduced as a positive control.The fluorescence signal was determined with a confocal laser scanning microscope (TCS SP8 SR Leica) at an excitation wavelength of 488 nm for GFP.

3.Results

3.1.Bioinformatic analysis of NIPs in cucumber,tomato and rice

Genomic analysis in cucumber showed that there are a total of nineCsNIPgenes,which are distributed on four chromosomes (Appendix A):CsNIP1;1was localized to chromosome 6;CsNIP1;2,CsNIP2;1,CsNIP2;2andCsNIP4;1were localized to chromosome 3;CsNIP3;1a,CsNIP3;1bandCsNIP3;2were localized to chromosome 5;andCsNIP3;3to chromosome 4. Synteny analysis showed thatCsNIP2;2is a tandem repeat ofCsNIP2;1(Appendix A),the encoding protein of which has demonstrated influx transport activity for silicon (Sunet al.2017),suggesting thatCsNIP2;2may also be a silicon influx transporter gene.

Phylogenetic analysis of NIPs in cucumber,tomato and rice showed that SlNIP4;3,SlNIP4;2,SlNIP4;1,CsNIP1;2,OsNIP1;2,OsNIP1;1,OsNIP1;4,OsNIP1;3,SlNIP1;2,SlNIP1;1,CsNIP1;1,SlNIP3;1 and SlNIP2;2 were clustered into one group (Fig.1-A). SlNIP6;1,CsNIP3;3,CsNIP3;1a,CsNIP3;1b,OsNIP3;1,SlNIP5;1,CsNIP3;2,OsNIP3;2 and OsNIP3;3 were clustered into another group. SlNIP2;1 and CsNIP2;1 have been identified as the silicon influx transporters Lsi1 in tomato and cucumber,respectively (Sunet al.2017,2020). OsNIP2;1 and OsNIP2;2 are both silicon influx transporters in rice,and named as OsLsi1 and OsLsi6,respectively (Maet al.2006;Yamaji and Ma 2009;Yamajiet al.2015). The phylogenetic analysis demonstrated that CsNIP2;2 was grouped with these four silicon transporters,and it was closest to SlNIP2;1 (Fig.1-A),a silicon influx transporter in tomato (SlLsi1;Sunet al.2020). This association again implies that CsNIP2;2 may be a silicon influx transporter.

Fig.1 Phylogenetic and structural analysis of nodulin 26-like intrinsic proteins (NIPs) in cucumber,tomato and rice. The evolutionary history of NIPs (A) was inferred by using the Maximum Likelihood method and Le_Gascuel_2008 model (Le and Gascuel 2008).The ar/R selectivity filter (B) and NPA/V motif (C) are shown on the right. Characterized silicon transporters and CsNIP2;2 are highlighted with an orange background.

In NIPs,four amino acid residues from the transmembrane helix 2,helix 5 and loop E form a selectivity filter to particular substrates (Murataet al.2000). Different groups of NIPs in cucumber,tomato and rice have distinctive selectivity filters. For instance,OsNIP1s (including OsNIP1;1/1;2/1;3/1;4) and SlNIP1s(including SlNIP1;1/1;2) share a selectivity filter of‘WVAR’;whereas the selectivity filter of SlNIP3;1 is ‘WIAR’(Fig.1-B). SlNIP2;1,CsNIP2;1,OsNIP2;1 and OsNIP2;2,which have been respectively characterized as silicon transporters in tomato,cucumber and rice (Maet al.2006;Yamaji and Ma 2009;Yamajiet al.2015;Sunet al.2017,2020),share a selectivity filter of glycine,serine,glycine and arginine (GSGR) (Fig.1-B). In cucumber,CsNIP2;2 has a selectivity filter of cystine,serine,glycine and arginine (CSGR),demonstrating a one amino acid residue distinction from the“GSGR”of the four known silicon transporters (Fig.1-B).

For the majority of NIPs (except SlNIP4;3) in tomato,rice and cucumber,two NPA/S/T/V motifs have been identified,mostly as NPA (Fig.1-C). For instance,CsNIP2;2 has two NPA motifs (Fig.1-C);while SlNIP1;1,SlNIP2;2,SlNIP6;1,CsNIP2;1 and CsNIP3;3 have one NPA motif and one NPS/T/V motif;whereas CsNIP3;1a,CsNIP3;1b,CsNIP3;2,OsNIP3;1 and SlNIP5;1 have one NPS motif and one NPV motif (Fig.1-C;Appendix B).

3.2.Isolation of CsNIP2;2

Genome annotation of cucumber ‘9930’ showed thatCsNIP2;2is 2 807-bp long with five exons and four introns,and it shares a similar gene structure withCsLsi1(Appendix C). In this study,the CDS ofCsNIP2;2was isolated and sequenced from cucumber ‘Mch-4’. The CDS showed 100% identity with that from cucumber ‘9930’(http://www.cucurbitgenomics.org/). CsNIP2;2 have two NPA motifs,and the ar/R selectivity filter is CSGR,which is different from the other NIP silicon transporters experimentally identified (Figs.1-B and C,2-A;Appendix D). The distinction in the selectivity filter was mainly due to a nucleotide difference:T inCsNIP2;2and G in the other silicon transporter genes (Fig.2-A).

Phylogenetic analysis of all NIP silicon transporters experimentally identified in plants showed that CsNIP2;2 was closer to GmNIP2-1,a silicon influx transporter in soybean (Deshmukhet al.2013). All the transporters were clustered into two groups in accordance with the monocotyledon and dicotyledon classification (Fig.2-B).

Fig.2 Conserved protein sequence alignment and phylogenetic analysis of all nodulin 26-like intrinsic protein (NIP) silicon transporters experimentally verified in higher plants. A,protein sequence alignment of silicon transporters belonging to NIPs in cucumber (Cucumis sativus L.),barley (Hordeum vulgare L.),rice (Oryza sativa L.),maize (Zea mays L.),pumpkin (Cucurbita moschat Duch.),soybean (Glycine max Merr.),tomato (Solanum lycopersicum L.) and wheat (Triticum aestivum L.). Nucleotide sequence around the second helix (H2) site is shown,and the H2 site and corresponding nuclear sequence are shown with a purple background. B,phylogenetic tree of the above silicon transporters and CsNIP2;2. The evolutionary history was inferred by using the Maximum Likelihood method and Le_Gascuel_2008 model (Le and Gascuel 2008).

3.3.Transport activity of CsNIP2;2

To investigate the transport activity of CsNIP2;2 for silicon,cRNA ofCsNIP2;2was injected intoXenopuslaevisoocytes and the influx transport activity for germanium (an analogue of silicon) was determined. The results showed that oocytes injected withCsNIP2;2cRNA absorbed more germanium than those injected with an equal amount of water,and the uptake was not altered under low temperature (Fig.3). These results indicate that CsNIP2;2 has an influx transport activity for silicon,and the transport is a passive process.

Fig.3 Silicon influx transport activity of CsNIP2;2 in Xenopus oocytes. Either CsNIP2;2 cRNA or water were injected into Xenopus oocytes. The oocytes were exposed to MBS containing 1 mmol L–1 Ge(OH),and the Ge concentration in the oocytes was determined after 30 min of incubation. Values are mean±SD (n=3). Different letters above bars indicate a significant difference at P<0.05

3.4.Promoter characteristics and expression pattern of CsNIP2;2

Cis-acting elements in the promoter region (1.5 kb upstream of the ATG site) ofCsNIP2;2were analyzed,and elements involved in anaerobic induction and light responsiveness were found (Fig.4-A;Appendix E),which suggests that the expression ofCsNIP2;2may be affected by an anaerobic environment and light. Relative qPCR was used to determine the expression pattern in cucumber seedlings in the absence or presence of exogenous silicon. The results showed thatCsNIP2;2was expressed in cucumber root,stem,lamina and petiole,and the expression in stem was about 2–3 times higher than in other samples. In the presence of exogenous silicon,the expression ofCsNIP2;2was significantly decreased in the stem,while it was not obviously altered in the other tissues (Fig.4-B),suggesting a tissue-dependent expression response ofCsNIP2;2to exogenous silicon.

Fig.4 Cis-acting elements and tissue expression pattern of CsNIP2;2,as well as its response to exogenously applied silicon. A,cis-acting elements in the promoter region (1.5 kb upstream of the ATG site). B,tissue expression pattern of CsNIP2;2 and its response to 1 mmol L–1 silicon treatment for 6 h. The internal control was the ubiquitin extension protein gene. Si+,addition of 1 mmol L–1 silicic acid;Si–,no added silicon. Values are mean±SD (n=3). Different letters above the bars indicate a significant difference at P<0.05.

3.5.Intracellular localization of CsNIP2;2

Transient expression of CsNIP2;2-eGFP fusion protein was introduced in onion epidermal cells to determine the intracellular localization of CsNIP2;2. The results showed that green fluorescence of the empty vector was seen throughout the epidermal cells,while fluorescent signals of the CsNIP2;2-eGFP fusion protein were detected in the cell nucleus,plasma membrane and an unknown structure inside the cell (Fig.5).

3.6.Prediction of additional silicon influx transporters in plants

With the amino acid sequences of all experimentally identified silicon influx transporters in higher plants (Appendix D),we searched the protein database in NCBI by BLAST and found another 186 putative silicon influx transporters (Appendix F).All of the NPA/V motifs and selectivity filters were analyzed,as shown in Fig.6. The results showed that although most of the predicted silicon influx transporters contain two conserved NPA motifs and an ar/R selectivity filter composed of GSGR,a few different motifs and selectivity filters were also found.For example,one of the motifs in XP_028125323.1 was NLA,which showed an amino acid distinction from the typical NPA/V motif. The selectivity filter was CSGR in NP_001292702.1(CsNIP2;2) and XP_008443412.1 (inCucumismelo,another cucurbitaceous plant);ASGR in XP_027928829.1,XP_017417585.1,XP_014496451.1,XP_007139574.1,TKY65183.1 and RDX57679.1;SSGR in AWS33565.1,GGGR in XP_028780466.1;-SGR in VAI60803.1;SSGR in XP_020697722.1,XP_020572807.1 and PWA79641.1;and GAGR in ONK73891.1. Each of these demonstrated a one amino acid residue distinction from the typical GSGR selectivity filter in the previously identified silicon transporters.

Fig.6 Phylogenetic tree,selectivity filters and NPA domains of predicted silicon influx transporters in higher plants. The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model (Jones et al. 1992). The NPA/V motif and ar/R selectivity filter are shown and any distinctions from the typical motif GSGR are marked with a red background. Silicon transporters that have been experimentally verified are labeled with“Si”.

4.Discussion

Plant NIPs have demonstrated transport activities for a wide range of solutes,including mineral elements and organic solutes (Mitani-Uenoet al.2011). However,most of the research has focused on model plants.Cucumber is an important vegetable crop,and the functions of NIPs in it are largely unclear. In this study,we isolated a cucumberNIPgene,CsNIP2;2,and found that it is a silicon transporter gene.

4.1.Transport activity and intracellular localization of CsNIP2;2 in cucumber

In this research,when analyzing NIPs in cucumber,we found that CsNIP2;2 is highly homologous to the identified silicon transporters in rice and tomato (Fig.1-A). This implies that CsNIP2;2 may be a silicon transporter,and this speculation was confirmed in theXenopus laevisoocyte uptake experiment,which showed that heterologous expression ofCsNIP2;2in the oocytes demonstrated significant influx transport activity for silicon (Fig.3). Moreover,the transport activity was not energy dependent (Fig.3),suggesting that CsNIP2;2 may transport silicon passively in cucumber,as has been found for Lsi1 and Lsi6 in rice (Maet al.2006;Yamajiet al.2008;Yamaji and Ma 2009),barley (Chibaet al.2009;Yamajiet al.2012) and maize (Mitaniet al.2009b).

In this study,the transient expression ofCsNIP2;2-eGFPin onion epidermal cells demonstrated that CsNIP2;2 was localized at the nucleus,plasma membrane and an unknown structure inside the cell (Fig.5). These results are inconsistent with those of previous studies.Zhuet al.(2019) found that this protein was localized to the endoplasmic reticulum (ER). However,we speculate that the actual localization they observed may be in the cytoskeleton,as the image that was presented was quite similar to that of the SlTRM5 localization result found in tomato (Wuet al.2018). Wanget al.(2015) found that CsNIP2;2 (named CSIT-2 in that paper) was localized to the plasma membrane when this gene was transiently expressed in tobacco leaves. The reason for these different results is unclear. It is also unclear whether the unknown structure we observed in this study was a part of the cytoskeleton,which remains to be investigated. In previous studies,the silicon influx transporters had been shown to be mostly localized to the plasma membrane,such as in rice (Maet al.2006),barley (Chibaet al.2009),maize (Mitaniet al.2009b),wheat (Montpetitet al.2012) and tomato (Sunet al.2020);whereas in pumpkin,Mitaniet al.(2011) found that the silicon influx transporter from the bloom genotype was localized to the plasma membrane,and that from the bloomless genotype was localized to the ER. These findings suggest that the localization of silicon influx transporters may be related to plant genotype. Whether this is the case in cucumber remains to be investigated by testing more genotypes in the future.

Fig.5 Subcellular localization of CsNIP2;2. A construct carrying CsNIP2;2-GFP was introduced into onion epidermal cells by gene gun transduction. GFP was used as a control (A–C).D,E and F show the localization of CsNIP2;2. A and D,GFP fluorescence;B and E,bright-field images;C and F,the merged images. Bar=75 μm.

4.2.Expression of CsNIP2;2 in cucumber

Cis-acting elements are conserved sequences in the promoter that may bind to specific transcription factors and further regulate gene expression (Wittkopp and Kalay 2012). In addition to transcription start and enhancer regions,elements involved in anaerobic induction and light response were found more than once in the promoter ofCsNIP2;2,which indicates that the expression ofCsNIP2;2may be regulated by an anaerobic environment or light. Previous research has shown that the expression of both silicon influx and efflux transporters are influenced by exogenous abscisic acid (ABA) and dehydration conditions (Yamaji and Ma 2007,2011). A drought related element was also found in the promoter ofCsNIP2;2(Fig.4-A),indicating that the expression ofCsNIP2;2may also be affected by water deficit stress.Othercis-acting elements related to endosperm-specific negative expression,meristem expression and gibberellin responsiveness were also found. However,the regulatory roles of thesecis-acting elements onCsNIP2;2expression remain to be investigated.

Expression analysis in our study showed thatCsNIP2;2was expressed in various cucumber tissues,including the root,stem and leaf (Fig.4-B),which is consistent with the results of previous research (Wanget al.2015). However,the tissue expression patterns of this gene were different in the two studies. In this study,in the absence of added silicon,the expression ofCsNIP2;2was obviously higher in the stem than in any of the other tissues including the root,lamina and petiole,in which the expression levels of this gene were at comparable levels (Fig.4-B);whereas Wanget al.(2015) found that the expression ofCsNIP2;2was much lower in the stem than in the root and leaf.It is unclear what caused the differences in tissue expression patterns. One possible explanation is that different cucumber cultivars were used,i.e.,this study used ‘Mch-4’,whereas the study by Wanget al.(2015)used ‘Zhonnong 8’.

Given thatCsNIP2;2is a silicon transporter gene,the response of its expression to exogenous silicon was investigated in different tissues. The results showed that the expression ofCsNIP2;2in the root and leaf was not obviously altered,but the expression in the stem was decreased (Fig.4-B). In a previous study,Wanget al.(2015) also investigated the expression ofCsNIP2;2(namedCSiT-2in that study) and observed an up-regulation of expression by silicon in the leaf and root. Different expression responses in the root of silicon transporter genes to added silicon have also been observed in other studies. In roots of rice and soybean,the expression levels of bothLsi1andLsi2were downregulated by exogenous silicon (Maet al.2006,2007;Deshmukhet al.2013). Similar responses were also observed in the expression ofLsi2in barley and maize(Mitaniet al.2009a). On the contrary,the expression ofStLsi1in potato andCsLsi2in cucumber were increased by silicon (Vulavalaet al.2016;Sunet al.2018). In addition,the expression ofHvLsi1in barley,ZmLsi1in maize andTaLsi1in wheat were not altered by added silicon (Chibaet al.2009;Mitaniet al.2009b;Montpetitet al.2012). Limited information is available on the expression responses of silicon transporter genes to silicon treatment in tissues other than roots. In potato,Vulavalaet al.(2016) reported that the expression ofStLsi1in the leaf was up-regulated,while the expression ofStLsi2was down-regulated in the tuber flesh but not altered in the leaf,root,skin or stolon. In cucumber ‘Mch-4’,the expression ofCsLsi1was decreased in the root and stem and increased in the lamina,but unchanged in the petiole (Sunet al.2017);whereas the expression ofCsLsi2was up-regulated in the root but not altered in the stem,lamina or petiole. These studies suggest that the differences in the expression responses of silicon transporter genes to silicon treatment may be related to plant genotypes and tissues. The mechanism for siliconinduced regulation remains to be investigated.

4.3.The ar/R selectivity filter of CsNlP2;2 and other putative silicon transporters in higher plants

All NIPs have two NPA motifs and an ar/R selectivity filter composed of four amino acids from the transmembrane H2,H5 and loop E.These four amino acid residues form the narrowest part of the channel pore,and are essential for the selective and distinctive transport functions (Fuet al.2000;Suiet al.2001). The two NPA motifs in previously isolated silicon influx transporters were NPA and NPV,and the selectivity filter was all composed of G,S,G and R (Fig.2-A). In this study,we found two NPA motifs in CsNIP2;2;and the selectivity filter is CSGR,the first amino acid of which are distinct from that of any previously identified silicon influx transporter (Fig.2-A).Despite of this distinction,CsNIP2;2 demonstrated silicon transport activity (Fig.3). Previously,Mitani-Uenoet al.(2011) found that amino acid substitution at the H2 site of OsLsi1 (G→A) did not affect the transport activity for silicon,boron or arsenic. However,Wallace and Roberts (2005) reported that the amino acid at the H2 site of NIP was critical for the transport of water,glycerol and formaldehyde. These results imply that the amino acid at the second helix position of the ar/R selection filter is not critical for the silicon transport activity,but may be critical for the transport of other substances.

Plant NIPs have evolved from bacterial arsenic efflux channels and the amino acid change in the selectivity filter has essentially contributed to their functional evolution (Pommerreniget al.2020). Further studies on the features of the selectivity filters may help to identify the transport functions of NIPs. In this study,we predicted 186 NIPs as putative silicon influx transporters(Fig.6). Among them,were 15 NIPs (including CsNIP2;2)whose ar/R selectivity filters were different from those of previously verified silicon transporters (Fig.2-A).Muskmelon NIP (XP_008443412.1 in NCBI) shares the same ar/R selectivity filter with CsNIP2;2 in cucumber.Some other putative NIP silicon transporters also had a distinction from the typical GSGR filter in the first amino acid of the selectivity filter. For instance,some NIPs had a selection filter of ASGR,such as XP_027928829.1 inVignaunguiculate,XP_017417585.1 inVignaangularis,XP_014496451.1 inVignaradiata,XP_007139574.1 inPhaseolusvulgaris,TKY65183.1 inSpatholobus suberectusand RDX57679.1 inMucunapruriens.AWS33565.1 inTrifoliumrepenshad a selectivity filter of SSGR,and VAI60803.1 inTriticumturgidumsubsp.duruof“-SGR”. XP_020697722.1 inDendrobium catenatum,XP_020572807.1 inPhalaenopsisequestrisand PWA79641.1 inArtemisiaannuahad filter of SSGR.Since the first amino acid of the ar/R selectivity filter is not critical for the silicon transport activity as discussed above,these NIPs may have a transport activity for silicon,although this still remains to be verified experimentally.In addition,ONK73891.1 inAsparagusofficinalisand XP_028780466.1 inProsopisalbahad selectivity filters of GAGR and GGGR,respectively (Fig.6),both of which showed a distinction from the typical GSGR filter in the second amino acid. It is unclear whether these two NIPs have a transport activity for silicon. Mitani-Uenoet al.(2011) found that the second amino acid substitution of isoleucine for serine in the selectivity filter resulted in a loss of silicon transport activity,and suggested that the amino acid at the H5 position plays a key role in silicon transport.

Among the putative NIP silicon transporters,a NIP inCamelliasinensis(XP_028125323.1 in NCBI) had a motif of NLA as well as NPA. Whether the substitution of P/L affects the substrate specificity of this NIP remains unknown. As implied by Mitani-Uenoet al.(2011),in addition to the NPA/V motifs and ar/R selectivity filter,other structural features may also be involved in the control of transport substrate specificity of NIPs,although these remain to be identified in future.

5.Conclusion

CsNIP2;2 is a silicon influx transporter for passive Si uptake in cucumber. It has a selectivity filter composed of cysteine,serine,glycine and arginine,and this silicon transporter was localized to the cell nucleus,plasma membrane and an unknown structure inside the cell. The subcellular localization and ar/R selectivity filter are different from those of the previously characterized silicon influx transporters in other plants. The findings may help us to better understand the functions of NIPs in cucumber plants.

Acknowledgements

This work was supported by the National Key Research and Development Program of China (2018YFD1000800)and the National Natural Science Foundation of China(32072561 and 31772290).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm