Application of CRISPR/Cas9 genome editing system for molecular breeding of orchids

Orchid is an important ornamental plant in Indonesia due to their natural beauty of flowers. In the tropical forest, orchids are being acquired for trading and commercial market. Thus, the effort is required to proliferate orchid in large quantities for conservation and improve the floral variation for plant breeding. The purpose of this study is to develop a firmed methodology of molecular breeding of orchids using CRISPR/Cas9 KO system. The plant material used was Phalaenopsis amabilis protocorms growth on NP medium+pepton (2 g/L). Protocorm were submerged in the culture of Agrobacterium tumefaciens that Ti‐plasmid had been filled with a T‐DNA construct of a pRGEB32 vector harboring sgRNA with PDS3 sequence. Detection for transformants was confirmed by PCR using HPT primers (545 bp), Cas9 primers (402 bp), PDS primers (280 bp) and trnL‐F (1200 bp) as an internal control. The results showed that 0.96% PDS transformants were obtained from PDS3T2 lines. Several transformant showed pale leaf color compared to non‐transformant plants. This study suggests that the target gene has successfully edited by CRISPR/Cas9 system and could be applied for that functional gene editing in orchids.


Introduction
Orchid (Familia: Orchidaceae) is an important horticul tural plant in Indonesia. Some orchids, such as Dendro bium and Phalaenopsis, are known as popular potted or namental plants for their beautiful and longlasting flow ers (Tong et al. 2019; Cai et al. 2015. On the contrary, the shooting emergences of orchids in term of seedling production are difficult to achieve naturally and the flow ering time is very time consuming. These conditions are common problems in orchid breeding. Therefore it is nec essary to overcome those limitations in order to promote and accelerate the seedling production and flowering time. Those problems could be solved through CRISPR/Cas9 genome editing system to edit shooting and flowering genes. CRISPR/Cas9 is the latest genome editing tech nology that has been already used in various animals and plants including model and nonmodel ones. With its high efficiency, CRISPR/Cas9 could be used for targeting biallelic genes simultaneously (Tsutsui and Higashiyama 2016).
Conventional genetic engineering strategy has several issues and limitations, one of which is the complexity as sociated with the manipulation of large genomes of higher plants. However, creation of novel tools for breeding and biotechnology, an application area of genetic engineering, has received significant focus resulting in accelerated de velopment of useful tools. At the present, gene editing techniques have the potential to substantially accelerate plant breeding. Genome editing is a group of technologies that give scientists the ability to change an organism's DNA (Hsu et al. 2014). Targeted genome editing using artificial nu cleases has the potential to accelerate basic research as well as plant breeding by providing the means to modify genomes rapidly in a precise and predictable manner. The Clustered Regularly Interspaced Short Palindromic Repeat (CRISPR)/CRISPRassociated protein 9 (Cas9) system is a new developed tool for the introduction of sitespecific doublestranded DNA breaks (Liu et al. 2017). Cas9 protein induces doublestrand breaks (DBS) (Shen et al. 2017), which then repaired by the intracellular system. These technologies allow genetic material to be added, re moved, or altered at particular locations in the genome, it can be edits the form of point mutation, insertion, and deletion through the nonhomologous endjoining (NHEJ) pathway (Chiruvella et al. 2013) or by gene replacement through homologydirected recombination (HDR) (Hahn et al. 2018).
This research objective is to develop CRISPR/Cas9 method for use in orchid in order to produce desirable al lele in less than 20 months compared to the conventional breeding that need several years. In this preliminary study, we used Phytoene desaturase (PDS3) gene which is a rate limiting enzyme in carotenoid synthesis. The PDS3 gene converts phytoene to colored ξcarotene in a twostep de saturation reaction (Wang et al. 2009). PDS gene encoding the phytoene desaturase enzyme is involved in carotenoid biosynthesis pathway (Naing et al. 2019), therefore muta tions in those genes cause albino phenotype (Koschmieder et al. 2017). In general PDS3 gene is widely used as a com mon marker gene in many plant species due to its ease of detection. Based on the easy detection of mutation pheno typically, this gene was chosen as target for genome edit ing. This is a preliminary study to find out how high mu tation frequencies can be obtained with the CRISPR/Cas9 genome editing system in P. amabilis orchids, which can be further applied to edit other functional genes in orchids.

Plant Materials
Mature P. amabilis capsules were sterilized with 75% ethyl alcohol, 0.1% (w/v) natrium hydrochloride, then washed with sterile distilled water. After sterilization, the capsule was opened by cutting it in half longitudinally, then the seeds were sown on NP + peptone media (2 g/L). Cultures were maintained in an incubator with a temper ature of day/night ±25°C, orchid seeds then grew into a protocorm (developing orchid embryo). In general, pro tocorms could be used as targets because of their high re generation capacity.

Vector Construction and Targets Selection
The vector pRGEB32 in Agrobacterium strain EHA105 was used to transfer the TDNA which contain CRISPR/Cas9 into orchid plant ( Figure 1). The TDNA region of this vector harbors hygromycin phosphotrans ferase (Hyg R ) as the selection marker genes. Hyg R gene was driven by the CaMV 35S (35S) promoter and ended by the 35S' terminator, and cloned into the TDNA in the pRGEB32 vector. Agrobacteriummediated transforma tion was conducted according the protocol described by Semiarti et al. (2010Semiarti et al. ( , 2011 for genetic transformation into plant. To create a genome edited P. amabilis by using CRISPR/Cas9 system, an RNAguided genome editing vector line pRGEB32 (Bsa I)Cas9 was used for express ing engineered sgRNA and Cas9 in cells. The sgRNA ex pression cassette was driven by the U3 promoter. This pro moter was a small nuclear RNA (snRNA) promoter that initiates transcription at an adenine nucleotide and drives the high expression of sgRNAs (Shan et al. 2013; Ma et al. 2015. Based on homology search by BLAST, we found mRNA sequence of PDS3 from Phalaenopsis equestris or chid (XM_020730635). It is 1767bp from AUG to UAG in length, and total length of mRNA in database is 2176 bp ( Figure 2A). The protein motive showed that there is Pf01593 amino oxidase protein family that might be im portant to function the phytoene desaturase enzyme in carotenoid biosynthesis pathway (Naing et al. 2019). The DNA sequences of P. amabilis were selected as the can didates to determine the NGG as the protospacer adjacent motifs (PAM). To study the efficiency of CRISPR/Cas9 genome editing in P. amabilis, we selected two targets in PHYTOENE DESATURASE 3 (PDS3) gene. PDS3 is in volved in carotenoid synthesis in wich the mutation shown albino phenotype (Qin et al. 2007). Twenty nucleotides were selected from the nucleotide sequence no 312346 region of the PDS3 genes. Target 1 (PDS3T1) and tar get 2 (PDS3T2) were separated by 9 bp. The sequence of PDS3T1 is TTCGTGACTGTTTACGTCCC and PDS3T2 sequence is CTAGAAGTGGTAATTGCTGG (Figure2B).
The specificity of these candidates are examined by BLAST search against the genome sequence of P. am abilis. We checked it by BLAST and found P. ama bilis PDS3mRNA is expected to be almost identical to that of P. equestris based on our sequence data ( Figure  2B). In order to make the CRISPR/Cas9 constructs, two DNA oligonucleotides were chemically synthesized for each gRNA. The two oligonucleotides were complemen tary for the corresponding sequence to the spacer, and the sequences GGC and AAAC were added to the 50 end of the forward and reverse oligonucleotides, respectively, to allow the formation of cohesive ends of BsaI restriction sites following annealing (Kui et al. 2017).

Selection of Transformant Protocorms and Plant Regeneration
After three days of cocultivation and transformation pro cess have done, the infected protocorms were transferred to selection medium containing 10 ppm hygromisin and incubated for 4 weeks. After that, survived protocorms that indicated transformant candidates with green color were transferred to NP medium with NAA: BA (1:2). Af ter protocorms grown on the medium, molecular analysis was carried out to see the integration of TDNA in plant genomes.

Genomic DNA Isolation
The leaf pieces (200 mg) were collected in a sterile 1.5 mL microfuge tube. Plant tissues were ground into a fine pow der, then added with 500 µL of 3% CTAB buffer. Plant extract mixture were incubated for 30 min at 60°C in wa ter bath. One volume of chloroform:isoamyl alcohol was added, mixed well and incubated for 30 min in room tem perature with shaking for homogenation, then centrifuged the mixture at 5000 rpm for 5 min. The upper phase so lution was carefully transferred to a new sterile 1.5 mL tube. The volume of supernatant was determined. One tenth volume of Na acetate or cold absolute EtOH was added and mixed gently, then incubated at -20°C for 10 min. Sample was centrifuged at 1000 rpm for 5 min and supernatant was discarded. The pellet was washed with 70% ethanol by centrifugation at 1000 rpm for 2 min. To dry up the DNA pellets, tubes with their lids open were inverted on sterile tissue paper for 30 min then DNA pel lets were dissolved in 2530 µL TE buffer/ Nuclease free water each. Isolated DNA samples were stored in -20°C . The genomic DNA concentration in resulting solution (200 µL) was measured by spectrophotometer (Nanodrop 200C, Thermoscientific).

PCR test for transformant candidates
In this study the direct PCR method was performed using KOD FxNeo PCR Kit (Toyobo, Japan). The DNA tem plate used was obtained by cutting a small protocorm (± 0.05 cm) of the plant transformant candidates using sterile tweezers embedded. Samples of plant parts were crushed by mixing 30 µL buffer solution (buffer solution A: 100 µM TrisHCl pH 9.5; 1 M KCl; 10 mM EDTA) with mi cropestle. DNA templates were obtained from the super natant. PCR was performed in a 25 µL volume contain ing 1 µL genomic DNA sample, 12.5 µL 2× PCR buffer for KOD Fx Neo, 5 µL 2 mM dNTPs, 0.5 µL KOD Fx Neo (Toyobo, Japan), 0.75 µL of each primers and 4.5 µL nucleasefree water. The PCR conditions were pre denaturation (94°C for 2 min), 30 cycles of denaturation (98°C for 10 s), annealing (58°C for 50 s), and exten sion (68°C for 45 s). The PCR was performed using T100 ™ Thermal Cycler (BioRad). The primer sequences are shown in Table 1.

Qualitative analysis of transformant
To find out whether there is a mutation at the target site in the transformant genome, a PCR with primers specific to the target gene was performed. Plant genome DNA were amplified by PCR using PDS primers and separated the PCR product using 4% low EEO agarose S (Nippon Gene, 31390231). The low EEO agarose was dissolved in 1× TrisAcetate EDTA (TAE) bufer (40 mM Trisacetate, 0.5 mM EDTA pH 8.3, Fisher Bioregents BP13324) by heating the solution in a microwave oven for 2-3 min. PCR with some primers: HPT, Cas9 and trnLF were also conducted using specific primers and checked us ing 2% agarose gel (Sigma Aldrich type 2). The melted agarose gels were added with 5 µL of ethidium bromide (10 mg/mL), then it was immediately poured on a UV transparent gel casting tray. High concentration of agarose gels should be poured rapidly as the gel solidifies quickly. The electrophoresis chamber filled with 1× TAE buffer till about 1 cm above the gel. The total of 10 µL of each PCR sample was loaded into each well and electrophoresis was performed for 1 h and 40 min. Our power supply was set to 100 V. The 4% agarose gels were run at room temper ature (25°C). DNA size marker 1kb (Geneaid) is a ready touse solution containing 6× loading dyes. Gel images were acquired using a regular geldocumentation system (Bhattacharya and Van Meir 2019).

Plant genetic transformation mediated by Agrobacterium tumefaciens
In this research protocorms were used for transforma tion material using Agrobacterium tumefaciens strain EHA105 carried TDNA with UBI::Cas9::U3::PDS in pRGEB32. The transformation efficiency in P. amabilis using PDS3T1 sgRNA is 0.9%, lower than transforma tion efficiency using PDS3T2 sgRNA is 0.96% (Table 2). Transformation efficiency in this study were lower com pared to the previous study. In Dendrobium phalaenopsis, transformation efficiency reached until 12% (Setiari et al. 2018) and D. lasianthera up to 70% (Utami et al. 2018). This might be caused by each species has their own ge netic make up (Men et al. 2003). Moreover, transforma tion by using Agrobacterium was a biological approach that has various factor to get an optimum result, such as type explant, bacteria condition and pretreatment (Chen et al. 2018). Agrobacteriummediated transformation is a method that enables production of stable transformants and as a model for studying the cellular localization and inter actions between proteins that are important for the de velopment of functional genomic analysis (Krenek et al. 2015). Agrobacteriummediated transformation used by scientists can introduce new traits into plants through over expression, knockdown or knockout the gene expression through Agrobacteriummediated TDNA transfer (Semi arti et al. 2007; Guo et al. 2019. A. tumefaciens uses a unique virulence mechanism to induce tumors, it delivers DNA fragments (transferred DNA or TDNA) to host cells where the TDNA ultimately integrates into the genome. The pathogenicity of A. tumefaciens is predominantly de termined by a large Ti (tumor inducing) plasmid (Guo et al. 2019).
The phenotypic change was also detected from the green color reduction of almost all protocorms, that changes from green to black on the 4th week in the selec tion medium. The black phenotype of protocorm can be caused by hygromycin as a selection agent affected pho tosynthesis rate. As MingeotLeclercq et al. (1999), this antibiotic can bind 80 ribosomal complexes. Furthermore, the green protocorm was indicated as a transformant can didate plant (Figure 3). In this result the transformation efficiency was low (Table 2), it might be due to the low efficiency in T DNA integration into the plant genome. Our previous study in Dendrobium phalaenopsis, efficiency transfor mation reached until 12% (Setiari et al. 2018) and Utami et al. (2018) got high transformation efficiency in D. lasianthera that the transformation efficiency up to 70%. This might be caused by specific character of each species related to their own genetic make up (Men et al. 2003). Moreover, transformation by using Agrobacterium was a biological approach that has various factor to get an op timum result, such as type of explant, bacteria condition and pretreatment (Chen et al. 2018). A. tumefaciens mediated genetic transformation has been routinely used in rice. The study used different cultivar rice with A. tume faciensmediated genetic transformation. However, the transformation efficiency of the indica rice variety is still much lower than that of japonica cultivars. This further improvement on the transformation efficiency lies in the genetic manipulation of the plant itself, which requires a better understanding of the underlying process in account for the susceptibility of plant cells to Agrobacterium infec tion as well as the identification of plant genes involved in the transformation process (Tie et al. 2012). Utami et al. (2018) used the five different infection period of Agrobac terium for another purpose. According to our result, futher research is necessary to increase mutation frequency for mutant propagation.

PCR-based detection of CRISPR/Cas9-edited in transformant plant
Transformant candidates were screened using PCR to con firm TDNA integration. HPT and Cas9 primers were used to screen the candidates to confirm the transformations of pRGEB32 construct harboring TDNA (UBI :: Cas9 :: U3 :: PDS). After that, 8 weeksold green protocorms grown on selection medium were used as screening materials. We regenerated the transformant in NP media+ ZPT (NAA: BA 1:2). Protocorms were grown in regeneration media until they produced leaves for mutation detection accord ing to Bhattacharya and Van Meir (2019) method. The electrophoresis results (Figure 4) show that T DNA has successfully entered the plant genome. This refers to the band that is detected appearing on some of the primers used. Results showed specific size of amplicon from each primer pairs, i.e HPT (504 bp), Cas9 (402 bp), PDS (280 bp) and trnLF (1200 bp) as an internal control. Results showed no multiband or band shift on gel elec trophoresis. This possibility does not occur in the genome insertion so that no extra thickening of the band or bands appear. So it needs to be confirmed by using Sanger Se quencing analysis method to see whether there are the mu tations occurred or not.

Sequence analysis of CRISPR/Cas9-edited in transformant plant
Mutation detection in the genomes of transformant plants were analysed by using 4% agarose gel low EEO following the protocol of Bhattacharya and Van Meir (2019). Nev ertheless, the drawback is that the amplicon size cannot be known accurately. It can only be estimated its size by com paring nontransformant with transformants and referring to the markers used. Further confirmation is required by sequencing and alignment to find specific changes in tar get sequence. In this research, we compared sequencing data from wild type/non transformant and transformant. We used PDS3 gene with two target site namely PDS3T1 and PDS3T2. Sequencing analysis uses PCR amplicons produced from genomic DNA templates. Furthermore, the process of alignment of the transformant genomic DNA template that is identical to wild type sequences. This analysis showed that one of the edited protocorms kept a muta tion ( Figure 5) located on the CRISPR/Cas9 target site for  PDS3T2. This indicated that mutations occured as a base deletions due to internal DNA repair, then the process of cutting double strand DNA and forming a doublestrand break were caused by CRISPR/Cas9. The repair process is called NHEJ (Nonhomologous end joining) gene mu tation that results in a knockout. This mutation induces changes in amino acid sequences which will also changes the phenotypic. Because of those carotenoid pathways dis rupted causing an albino phenotype ( Figure 6). In our pre vious study, to edit the homologous PDS3 gene in P. ama bilis orchids, we used the pKIR1.1 vector (Tsutsui and Hi gashiyama 2016) with the AtRPS5A promoter from Ara bidopsis, but we did not get any mutants that showed in sertion or deletion at the target site of PDS3 gene. The pRGEB32 vector that containing OsUbipro, a Ubiqui tin promoter from rice (Oryza sativa) which is a mono cot plant similar to an orchid, turns out that CRISPR/Cas9 can work properly in orchid, produced orchid mutant with color changes on the leaves because of the indel occu rances in their PDS3 target sites.
Our in silico sequence allignment of the wild type and CRISPR/Cas9edited P. amabilis T#6 PDS3T2 (Fig  ure 7) showed that deletion changed the translated amino acid sequence. Indicated by black triangles. Yellow highlight was PAM sequence. This data indicated that CRISPR/Cas9 can induced a mutation in the way of dele tion. This phenomenon lead a frameshift on the translated amino acid sequence.
This shows that the CRISPR/Cas9 method can be used to edit orchid genome using appropriate vectors and pro

Conclusions
CRISPR/Cas9 genome editing KO system can be applied to orchids, especially Phalaenopsis amabilis. The PHY TOENE DESATURASE 3 (PDS3) gene can be used as a model/marker gene which makes it easy to determine whether the two PDS3 alleles are disturbed in the observed tissue. The pds3 mutants of P. amabilis show albino phe notype in the leaf tissues. The use of CRISPR/Cas9 for P. amabilis is still in its early stages in the T0 generation, so further research is needed for the the next generation in order to obtain precise information about the effective ness next generation of the editing genome in orchids, es pecially for KO suppressor functional genes in producing new trait.
settes. SN, YS and MDL conducted the genetic transfor mation bacteria and screening in antibiotic supplemented plates. SN conducted genetic transformation and YS con ducted the isolation of genomic DNA, PCR. SN con ducted, sequencing, subsequent analyzes, data compiling and wrote the manuscript. All authors read and approved the final version of the manuscript.