Introduction

Oil palm is the primary commodity crop in Malaysia, with oil palm planted area in 2022 reaching 5.67 million hectares (Parveez et al. 2023). This golden crop belongs to the family Arecaceae (formally known as Palmeae) of the monocot order Arecales and the genus Elaeis (Ong et al. 2020). There are two species of Elaeis, namely, E. guineensis Jacq. and E. oleifera, originated from West Africa and Latin America, respectively. Although both species can produce edible oil from fruit mesocarp and kernel, only the African oil palm E. guineensis is commercially planted due to its high oil yield. The oil yield per hectare of commercial oil palm is ten times higher than any other oilseed crop, making it the world's most efficient and valuable oil crop (Parveez et al. 2020). The demand for fats and oils derived from oil palm is increasing yearly due to the growing global population. Therefore, improving oil yield is necessary to meet food and industrial requirements. Oil palm improvement through breeding techniques is the key to increasing the oil yield without using more land (Parveez et al. 2015). However, oil palm breeding has a few limitations, such as a narrow genetic base and a highly heterozygous nature, because it is cross-pollinated, perennial, and requires heavy resources (land and other inputs). Therefore, the genetic engineering approach is one of the possible solutions to produce oil palm with high oil productivity, high oil quality, and other desired novel traits, thus ensuring the sustainability of the oil palm industry.

In oil palm genetic engineering, various works have been extensively conducted, such as the identification, characterization, and isolation of promoters and functional genes from oil palm, the construction of numerous transformation vectors, and the optimization of various oil palm transformation methods (Masani et al. 2018, 2022; Fizree et al. 2023). The oil palm genetic transformation was initially started using a biolistic method followed by Agrobacterium, polyethylene-glycol (PEG)-mediated transformation, and DNA microinjection (Parveez et al. 1998; Masli et al. 2009; Masani et al. 2014). Among the target goals of oil palm genetic engineering are producing high oleic acid, high palmitoleic acid, high lycopene, high ricinoleic acid, and biodegradable plastics (Parveez et al. 2015; Masani et al. 2018).

Numerous molecular approaches have been explored to induce mutations in plant species, such as gene silencing by RNA interference; however, this method does not allow for targeted genome editing. Novel genome-editing alternatives known as zinc finger nucleases (ZFNs) and TAL effector nucleases (TALENs) have been discovered (Iqbal et al. 2020). ZFNs and TALENs allow for precise nucleotide modifications of a gene of interest. However, because protein engineering is necessary for editing the gene of interest, implementing both technologies has been time-consuming and expensive (Gaj et al. 2013; Bortesi and Fischer 2015). In 2013, a simple yet versatile solution, the clustered regularly interspaced short palindromic repeat (CRISPR)-associated protein9 (Cas9) or CRISPR/Cas9, for targeted gene editing was discovered (Doudna and Charpentier 2014). Compared to ZFNs and TALENs, the CRISPR/Cas9 system is more practical and cost-effective.

The CRISPR/Cas9 system is a versatile genome-editing technology that uses short RNAs as guides or guide RNA (gRNA) to target multiple genes (Doudna and Charpentier 2014). It was initially discovered as an essential element in Escherichia coli in the 1980s for archaeal and bacterial adaptive immunity (Ishino et al. 1987). However, until it was revealed that Streptococcus thermophilus could develop resistance to a bacteriophage by integrating a genetic segment of an infectious virus into its CRISPR locus, the system's extraordinary function was unknown (Barrangou et al. 2007). This approach relies on a double-stranded break (DSB) driven by a CRISPR RNA (crRNA) transcript to disrupt the target DNA (Bortesi and Fischer 2015). Two forms of DNA repair emerge after DSB of targeted DNA, known as non-homologous end-joining (NHEJ) and homology-directed repair (HDR) pathways. NHEJ is an error-prone mechanism that thus results in poor repair, causing gene function to be disrupted. On the other hand, HDR utilizes a template to repair DNA and makes new DNA complementary to the repair template (Belhaj et al. 2015). Thus far, numerous gene-editing types of research have been accomplished in over 70 different crop species (Timofejeva and Singh 2023), including monocotyledons such as sorghum (Jiang et al. 2013), rice (Shan et al. 2014), maize (Sant’Ana et al. 2020), soybean (Kim and Choi 2021), and dicotyledon plants, such as Populus (Fan et al. 2015), tomato (Nicolia et al. 2021), and pea (Li et al. 2023).

Enhancing plant qualities has always been the prime objective in agriculture genome editing efforts. Although there is a vast array of plant species, the breeding objectives remain comparable. For instance, amplifying the concentrations of select unique secondary metabolites, lengthening the preservation period of fruits, modifying the growth pattern of trees, optimizing yield potential, and fortifying resistance to plant pests and diseases. Through genome editing, these desired traits could be generated by modifying the related genes. For example, the use of CRISPR/Cas9 genome editing has led to the achievement of producing wheat plants with resistance to powdery mildew disease, a condition caused by the fungal pathogen Blumeria graminis (Wang et al. 2014). The utilization of CRISPR/Cas9 technology enables researchers to effectively enhance disease resistance in cucumber (Cucumis sativus) by precisely targeting the gene that encodes translational initiation factor eIF4E (Chandrasekaran et al. 2016). This precision editing results in the development of cucumber plants that possess robust and versatile immunity against various viruses.

Genome editing via CRISPR/Cas9 has also been used to increase crop productivity. For example, CRISPR/Cas9 has been used to simultaneously knock out three major genes that are responsible for regulating the size of the rice grains. The generated plants produced an increase in grain size and weight by up to 20%-30% when compared with the wild type (Xu et al. 2016). Similarly, the CRISPR/Cas9 system was also used to knock out the oleoyl-CoA desaturase (FAD2) gene to produce Camelina sativa plants with an increased oleic acid from 16 to 50%, and concurrently decreased linoleic and linolenic acids to least than 4% and 10%, respectively (Jiang et al. 2017). In addition, in weed control management, CRISPR/Cas9 has been utilized to develop plants that possess tolerance to herbicides. For example, CRISPR/Cas9 targeting the phosphoenolpyruvate-binding site in flax (Linum usitatissimum) produced edited flax lines that are tolerant to glyphosate (Sauer et al. 2016).

In 2013, the publication of the whole-genome sequence of African oil palm species (Singh et al. 2013) provided a wide-ranging chance to transition towards oil palm genome editing. The first report on successful genome editing in oil palm was published by Yeap et al. (2021), targeting the phytoene desaturase (EgPDS) and brassinosteroid-insensitive 1 (EgBRI1) genes. Based on the report, 62.5% to 83.33% mutation efficiencies were obtained using the biolistic transformation of oil palm immature embryos with CRISPR/Cas9 vectors. The report demonstrated that the CRISPR/Cas9 gene-editing system could induce site-targeted mutagenesis in the oil palm genome. The successful adaptation of the CRISPR/Cas9 system was later reported by Bahariah et al. (2023), targeting the genes responsible for increasing oleic acid content in oil palm. Later, the strategy in designing and selecting efficient gRNAs for oil palm study was reported by Jamaludin et al. (2023), marking a closer step to using CRISPR/Cas9 in oil palm genetic engineering.

Even though attempts are made towards an efficient CRISPR/Cas9 system, various plant species may differ in terms of their efficiencies and challenges (Cardi et al. 2023). The difficulties in oil palm transformation due to the size and complexity of genomic may raise the risk of off-target mutations and reduce gene-editing specificity. The delivery of CRISPR/Cas9 reagents into oil palm tissues could be further hampered by its poor transformation efficiency coupled with long plant generation time. Therefore, developing strategies that ensure efficient expression of large numbers of different gRNAs simultaneously from a single vector would allow more extensive use of the multiplex capability of the CRISPR/Cas9 system in oil palm. Implementation from available CRISPR/Cas9 protocols that have been developed in other evolutionary closely related species, such as rice, can also be adopted. Applying a DNA-free CRISPR/Cas9 system via ribonucleoprotein (RNP) in oil palm can avoid the red tape involving genetically modified organism (GMO) regulations. Therefore, this review mainly discusses the strategies and applications of CRISPR/Cas9 genome editing in plants to identify the workable CRISPR/Cas9 system for oil palm that could also lead to DNA-free gene-edited oil palm for a sustainable future.

Discovery of CRISPR/Cas9 genome editing system

Three decades ago, in 1987, a group of researchers in Japan discovered unknown DNA sequences made up of short repeating nucleotides flanked by a short segment (Ishino et al. 1987). In 2002, the strange bacterial sequence was identified as clustered regularly interspaced short palindromic repairs (CRISPR) and CRISPR-associated protein 9 (Cas9) or CRISPR/Cas9, a versatile, highly potential tool for genome editing (Jansen et al. 2002). Using whole-genome sequencing (WGS) technology to study CRISPR revealed that this system was commonly found in viruses and plasmids, thus possibly encoding for bacterial immunity system (Jansen et al. 2002). Consequent biochemical and genetic studies validated the previous speculation, demonstrating that the CRISPR/Cas9 system is involved in detecting and protecting mobile genetic elements (Barrangou et al. 2007; Charpentier and Doudna 2013).

The Cas9 gene was later discovered to work with the gRNA to locate and cut the invading DNA at the conserved region, also known as the proto-spacer adjacent motifs (PAM) sequence (Carroll 2012). Further research discovered that crRNA and trans-activating CRISPR RNA (tracrRNA) were required for Cas9 to form a functional complex for CRISPR/Cas9 activity (Deltcheva et al. 2011). Intriguingly, this dual CRISPR RNA and trans-activating CRISPR RNA can be recreated as gRNA that is enough to drive Cas9 to induce DSB in target DNA (Jinek et al. 2012). The true potential of genome editing applications using the CRISPR/Cas9 system fully emerged when research revealed that the Cas9 guided by the gRNA could work in various types of cells and organisms and cut the genome with high specificity (Charpentier and Doudna 2013). Subsequently, when DSB occurred due to the action of Cas9, two DNA repair mechanisms, NHEJ or HDR, were activated. Both pathways will result in DNA modification at the repair or target sites. Fast-forward to the discovery of the CRISPR/Cas9 system, this technology has been advanced vigorously, focussing on efficiency, enhancing specificity, and expanding its application in various organisms, including plants.

CRISPR/Cas9 versatility, architecture, and mechanisms

The Streptococcus pyogenes Cas9 or SpCas9 gene is a 4107 bp full-length coding region originating from the type II prokaryotic CRISPR adaptive immune system (Wu et al. 2014). It consists of two separate nuclease domains, namely RuvC and HNH domains. Both can cleave one of the target DNA strands and simultaneously generate a blunt-ended DSB. The gRNA is a fusion of crRNA and tracrRNA. It is crucial for CRISPR activity to recognize and cleave the target genomic sequence and maintain stability, thus improving the CRISPR/Cas9 system (Okada et al. 2022). The target recognition sequence, including PAM, will be bound with the Cas9 protein to form a Cas9/gRNA complex. This complex specificity towards its target site vitally depends on the hybridization of the target recognition sequence (without PAM) of gRNA to that specific target site. During the process, the PAM sequence acts as the binding signal for Cas9. The target sequence will initially recognize and move towards its NGG PAM sequence, allowing the Cas9/gRNA complex to locate its target DNA site (Barman et al. 2020). The complex will then separate the target DNA strands to facilitate the complementary base-pairing of gRNA with the target DNA strand. The RuvC and HNH domains will then cut off one of the two strands at three bases upstream of PAM, subsequently generating blunt-ended DSB. The DSB will be repaired either by an error-prone mechanism, so-called non-homologous end-joining (NHEJ), or an error-free mechanism, namely a homology-directed repair pathway (HDR) (Fig. 1). Both pathways can be used to induce gene modifications at the target loci. However, the choice of repair mechanism depends on various factors, including the stage of the cell cycle, the nature of DSB, and the availability of the repair templates (Ceccaldi et al. 2016).

Fig. 1
figure 1

Schematic illustration of CRISPR/Cas9 genome editing mechanisms. The Cas9–gRNA complex binds any genomic region with a PAM sequence (shown in pink). The 20 nucleotides of the gRNA sequence (without PAM) complement the genomic sequence immediately upstream of PAM (red). The Cas9 endonuclease will cleave the DNA genomic region at three nucleotides upstream of PAM (indicated by scissor), introducing the double-strand break (DSB). The DSB will be repaired by either an error-prone NHEJ pathway or HDR, an error-free mechanism. The target gene would be knockout if the indels were introduced at the exon or gene promoter. In contrast, if the HDR pathway repairs the DSB, gene replacement will be introduced to the genomic region adjacent to the DSB

Full size image

Selection of the target gene for the establishment of the CRISPR/Cas9 system

Selection of the target gene is based on a case-by-case basis, mainly depending on the objectives of the research and the characteristics of the target gene, including the mutant’s phenotyping efficiency and the gene copy number (Shan et al. 2020). The target gene sequence and genomic structure identification, such as intron–exon structure in the target plants, can be obtained from a BLAST search using the available transcriptomic data (Shan et al. 2018). Further sequence confirmation and additional information on the corresponding genomic sequence can be obtained using PCR amplification and sequencing of the target amplicons. This information will facilitate the identification and selection of suitable gRNA. For an initial demonstration of CRISPR/Cas9 principles, loss-of-function gene activity is commonly being used by targeting the gene that produces a mutant with an obvious change in the phenotype, i.e., knockout of phytoene desaturase (PDS) that produces mutants with the albino phenotype (Charrier et al. 2019; Dai et al. 2021) (Table 1). This type of study has been applied as the first CRISPR/Cas9 application in various plant species, including oil palm (Yeap et al. 2021; Jamaludin et al. 2023), because it is more straightforward than targeting the gene controlling the complex traits.

Table 1 Recent applications of CRISPR/Cas9 system in oil palm and other plant species
Full size table

Furthermore, the efficiency of CRISPR/Cas9 genome editing can also be evaluated by inducing mutation in the exogenous gene (DNA or gene originating from outside of the studied organism, also known as transgene) in the transgenic mutant (Kishi-Kaboshi et al. 2017). For example, Bhowmik et al. (2018) have designed gRNAs to target the red fluorescent protein gene from Discosoma coral (DsRED) in transgenic wheat microspores. With the availability of the DsRED sequences, the gRNA sequence can be efficiently designed. The absence of the red fluorescence signals indicated the successful knockout of the DsRED gene by CRISPR/Cas9 in transgenic wheat microspores. For oil palm, a series of DsRED gene constructs was developed and successfully tested in oil palm calli as a visual reporter marker (Fizree et al. 2019) and could be potentially used to evaluate CRISPR/Cas9 genome editing in oil palm.

Parameters for an efficient gRNA design

In CRISPR/Cas9 genome editing, designing a gRNA for gene targeting is the first step in CRISPR/Cas9 experiments (Doench et al. 2014). The sequence of gRNA can significantly affect the target DNA's cleavage efficiency, which also influences the probability of off-target binding and Cas9 cleavage (Zischewski et al. 2017). Therefore, designing the correct gRNA is critical for the success of CRISPR/Cas9 experiments. There are several important parameters to consider when designing a gRNA. The binding of each Cas nuclease to the target sequence happens only in the presence of a PAM sequence (Wu et al. 2014). For that reason, the target sites in the genome can be targeted by different Cas nuclease proteins that are restricted by the positions of particular PAM sequences (Molla and Yang 2020). This is because each Cas nuclease isolated from different bacterial species recognizes a different sequence of PAM (Kleinstiver et al. 2015). For example, SpCas9 nuclease cleaves the nucleotide upstream of the PAM sequence 5'-NGG-3' (where "N" can be any nucleotide base), while PAM sequence 5'-NNGRR(N)-3' is required by SaCas9 (Staphylococcus aureus) to target DNA regions for CRISPR editing (Ran et al. 2015). Although the PAM sequence itself plays an essential role in the cleavage of target sequences, it must not be included in the sequence of a gRNA.

Doench et al. (2014), Wu et al. (2014), Wong et al. (2015), Liang et al. (2016), and Bruegmann et al. (2019) reported that the presence of specific motifs in the gRNA sequence and gRNA secondary structure could enhance the activity of the gRNA in CRISPR/Cas9 studies (Fig. 2). For example, there should be favorable nucleotides at several positions. Guanine was preferred directly after the PAM motif (position 20) but unfavored at position 16 (Doench et al. 2014). In addition, cytosine was unfavorable at positions 3 and 20 of the gRNA sequence but favored at position 16 over guanine (Wu et al. 2014). Furthermore, adenine was also preferred in the middle of the gRNA sequence (Doench et al. 2014), and purine residue was preferred in the last gRNA nucleotide sequence (Bruegmann et al. 2019). On the other hand, all gRNAs having TTTT (poly-T) were also disfavoured, because a stretch of identical contiguous nucleotides will be recognized by U6 promoters and terminate the RNA polymerase III transcription (Wong et al. 2015). An early study on non-functional gRNA motifs also suggested that repetitive uracil (UUU) in the seed region of gRNA (10–12 nucleotides before the PAM sequence) or at the last six bases of the gRNA could also impair the CRISPR/Cas9 activity (Wu et al. 2014). Other than repetitive UUU, four repetitive G nucleotides were also disfavoured, because it was reported that the GGGG could significantly lead to low CRISPR/Cas9 efficiency. This GGGG was also prone to forming a guanine tetrad. This secondary structure causes the gRNA sequence to be less accessible for gRNA target recognition and binding (Wong et al. 2015).

Fig. 2
figure 2

The characteristics of efficient gRNAs based on the 20-bp guide sequence and gRNA secondary structure analysis. These characteristics were based on research of highly active gRNAs reported by Doench et al. (2014), Wu et al. (2014), Wong et al. (2015), Liang et al. (2016), and Bruegmann et al. (2019)

Full size image

Another critical aspect of the high activity of gRNA is the secondary structure, which also affects the efficiency of the designed gRNA. Extensive studies reported that some structural motifs were identified to influence the effectiveness of gRNA positively (Doench et al. 2014; Wong et al. 2015). Although the current gRNA structure is based on bacterial crRNA and tracrRNA, which have been broadly used in eukaryotes, it has been shown that the gRNA secondary structure is prone to structural changes that may affect the CRISPR/Cas9 activity. Therefore, gRNA secondary structures should be analyzed to select the efficient gRNA that can produce high editing efficiency using the CRISPR/Cas9 system. The secondary structure analysis of designed gRNA can be done using the bioinformatics tool such as RNA fold Webserver by applying the Zuker–Stiegler algorithms and Andronescu model based on minimum free energy (MFE) representation (Bruegmann et al. 2019).

Typically, the secondary structure of a gRNA includes a crRNA sequence that encompasses 20 nucleotides of gRNA and a repeat region comprising 12 nucleotides, accompanied by a tracrRNA sequence obtained from a combination of an anti-repeat region and three stem-loop structures (Liang et al. 2016). By linking together, the artificial tetraloop structures form a single RNA transcript structure. The repeat and anti-repeat section of the tetraloop are commonly called the RAR stem-loop (GAAA), stem-loop 1 (CUAG), stem-loop 2 (GAAA), and stem-loop 3(AGU) (Jamaludin et al. 2023). Doench et al. (2014) and Wong et al. (2015) reported that the last three base pairs of the gRNA (seed region), 18–20 nucleotide position of the gRNA sequence, and 51–53 nucleotide position on the secondary structure should be unpaired and freely accessible to allow the binding of Cas9 to the target sequence.

The base-pairing of the guide sequence with its target DNA can be impacted by a stable complex between the gRNA sequence and other nucleotide bases (Liang et al. 2016). There are three types of base-pairing scores, namely, the total base-pairing scores (TBP), consecutive base-pairing scores (CBP), and internal base-pairing scores (IBP). The TBP score measures the total base pairs in the gRNA guide sequence combined with other sequences, whereas the CBP score represents the count of bases in the guide sequence combined with other sequences. Meanwhile, the internal base pair in the gRNA is indicated by the IBP score. To ensure efficiency in selecting gRNA, Liang et al. (2016) emphasized the importance of considering sequences with less than 12 TBP scores, a maximum of 7 CBP, and 6 IBP.

Based on these guidelines, Bahariah et al. (2023) have designed four gRNA for targeting palmitoyl-ACP-thioesterase (EgPAT) and oleoyl-CoA desaturase (EgFAD2) genes in an attempt to alter these genes for potential increase in oil palm oleic acid content. The study revealed that gRNA with high GC content of 60% and 65% could induce 100% indels rate. By considering these crucial motifs and preferable secondary gRNA structure, seven efficient gRNAs were selected from 167 gRNAs predicted by CRISPR-GE software to completely down-regulate the expression of oil palm phytoene desaturase (EgPDS) (Jamaludin et al. 2023). These seven gRNA sequences consisted of specific motifs such as adenine at position 9–11, no cytosine at position 3, and purine residue in the last 4 nucleotides, with the secondary structure having stem-loop 2 and stem-loop 3. The In vitro cleavage assay demonstrated that these seven sgRNA were efficient for cleavaging the corresponding target regions of the EgPDS gene.

Type of vectors for CRISPR/Cas9 system

Besides selecting the high-efficiency gRNA, the type of CRISPR/Cas9 vector system used can also affect the CRISPR/Cas9 editing efficiency. The efficiency of the CRISPR/Cas9 system varies amongst species, depending on the regulatory elements used, such as the species-based codon-optimized Cas9, promoters, and terminators for regulating the expression of Cas9 gene and gRNA sequence. An efficient vector system is generally required to produce a high expression level of gRNA and Cas9 genes, particularly for oil palm that is recalcitrant to genetic transformation. In the CRISPR/Cas9 vector system, two different RNA polymerase systems are employed; RNA polymerase II (Pol II) is responsible for driving the expression of the Cas9 gene, while RNA polymerase III (Pol III) controls the expression of gRNA. RNA polymerase III promoters, such as U3 or U6 small nuclear RNA (snRNA) gene promoters, are typically utilized to transcribe gRNAs in cells (Zhang et al. 2017) (Table 1). To initiate transcription, these Pol III promoters depend on a highly specific 5′ nucleotide, where the U6 promoter necessitates 5′-Guanine (G) and the U3 promoter demands 5′-Adenine (A) (Jiang et al. 2013). It is usually followed by a poly ‘T’ (five-to-eight Ts) that functions as a transcription termination signal (Kor et al. 2023). Thus, enhanced specificity can be achieved by including a specific nucleotide at the 5′ end of the gRNA sequences.

The OsU3 and OsU6 promoters from Oryza are commonly used for the monocots plants (Zafar et al. 2020; Bahariah et al. 2021, 2023; Tripathi et al. 2021; Yeap et al. 2021). Meanwhile, AtU6 and AtU3 promoters isolated from Arabidopsis thaliana are broadly used for the dicot plant (Jang et al. 2021; Pavese et al. 2021; Zhang et al. 2021a; Ly et al. 2023). The U6 promoters from Arabidopsis, Oryza, and Camelina were more efficient than their U3 counterparts (Zhou et al. 2023). In wheat (Triticum aestivum L.), the TaU3 promoter exhibited higher efficiency compared to the TaU6, OsU3, and OsU6 promoters (Zhang et al. 2021b; Kor et al. 2023). In barley (Hordeum vulgare), the HvU3 promoter showed higher activity than the OsU3 promoter (Lee et al. 2021). Therefore, it is recommended to isolate U3 and U6 from oil palm and experimentally evaluate the compatibility and efficiency of these promoters to regulate gRNA before further analysis is made.

Nevertheless, it is noteworthy that these U6 and U3 promoters may not always work for all targeted genes because of the absence of spatiotemporal specific control, which is ubiquitously present and expressed in all cells, tissues, and at all stages of plant growth and development (Xie and Yang 2013). For cloning multiple gRNA into the vector, different U3 or U6 promoters are recommended to regulate the gRNAs. This minimizes the risk of transgene silencing (Ma et al. 2015). Otherwise, the polycistronic tRNA–gRNA strategy can be utilized (Xie et al. 2015). In addition, for the gRNA scaffold, the long version of the gRNA scaffold (76 nucleotides) was shown to have higher efficiency compared to the short version (42 nucleotides) (Hsu et al. 2013). Both scaffolds are commonly used in most CRISPR/Cas9 studies in plants, because they are easier to design and show higher efficiency than the dual crRNA:tracrRNA system (Miao et al. 2013).

Aside from that, for Cas9, various plant constitutive promoters were utilized (Ma et al. 2016). This includes the CaMV35S promoter from the virus and Ubiquitin promoters from several plant species, such as maize, rice, and Arabidopsis (Zhang and Showalter 2020). It was reported that the maize ubiquitin promoter was more efficient for CRISPR/Cas9 editing in both monocot and dicot plants than the CaMV35S promoter from cauliflower mosaic virus (Feng et al. 2018). One factor contributing to this is the CaMV35S promoters reduced activity in embryogenic cells and vulnerability to transgene silencing (Jiang et al. 2014).

Nevertheless, the CaMV35S and maize Ubiquitin promoters have been used for regulating the Cas9 gene due to their ability to drive high transgene expression in oil palm cells efficiently (Yeap et al. 2021; Bahariah et al. 2023; Jamaludin et al. 2023). It also recommended employing oil palm endogenous constitutive promoters for driving the Cas9 gene, such as ubiquitin extension protein 2 (UEP2), which was demonstrated to regulate the transgene in oil palm tissues well (Fizree et al. 2019; Masura et al. 2019). Other promoters such as Arabidopsis egg-cell (E.C) specific promoters, including E.C 1.1 and E.C 1.2 (DD45), pollen-specific promoters such as LAT52, sporogenous cell-specific promoters such as SPL, and the Yao promoter were utilized to express the Cas9 gene in the early developmental stages, thus, increase the chance of obtaining homozygous and heritable mutations (Zhang and Showalter 2020).

To improve Cas9 expression in plants, the Cas9 gene was also modified with plant codons by attaching the nuclear localization signal (NLS) at both ends of Cas9 for gene editing in various crops, including oil palm (Yeap et al. 2021). It was reported that up to 83.33% mutation efficiency was achieved when oil palm codon-optimized Cas9 gene coupled with gRNA targeting EgPDS gene were transformed in oil palm immature embryos. However, no comparison was made to determine the editing efficiency between the SpCas9 gene and the oil palm codon-optimized Cas9 gene. The NLS is commonly used to transport the Cas9 gene into the nucleus of eukaryotes. According to reports, the codon-optimized Cas9 gene resulted in greater editing effectiveness in various plant species when compared to the original Cas9 gene (Jiang et al. 2013; Nekrasov et al. 2013; Ma et al. 2015). For instance, the codon-optimized Cas9 gene consisting of 50–400 bp GC-rich sequence, mimicking the Gramineae genes, was proven to be highly efficient in rice (Ma et al. 2015).

Multiplex CRISPR/Cas9 genome editing system

Previous CRISPR/Cas9 systems in plant protoplasts have shown the utilization of a single guide, with mutation frequencies of 0.1% in grape, 6.9% in apple, 19% in rice, and 44% in tobacco (Woo et al. 2015; Malnoy et al. 2016). This approach depends on a single gRNA to target a gene of interest by inducing a single DSB at the targeted locus and relies on the NHEJ to cause a frameshift mutation. However, this gene-knockout study is predicted to be less efficient in oil palm. Because oil palm is a diploid crop species with low transformation efficiency, multiple gRNAs may be required to edit the oil palm gene efficiently. Multiplex genome editing system consists of multiple gRNAs with distinct target sequences that bind to individual Cas9 to form the Cas9/gRNA complexes and subsequently bind and cleave the target genes in the same cell simultaneously (Cong et al. 2013).

In early works of CRISPR/Cas9 in plants, gRNA delivery was performed by co-transformation using several plasmids, such as in Arabidopsis, rice, and tomato (Timofejeva and Singh 2023). The method is tedious, even though multiple-site mutations were successfully achieved in these species. This type of strategy also can decrease the effectiveness of CRISPR/Cas9 for multiple gene editing in plant genomes. Therefore, various vector systems with multiplexing abilities were introduced. For example, the pYLCRISPRCas9 vector system designed by Ma et al. (2015) allows the assembly of multiple gRNA expression cassettes into a single vector system using golden gate cloning or the Gibson assembly method. This vector system is being used to develop a CRISPR/Cas9 system in oil palm (Bahariah et al. 2023; Jamaludin et al. 2023). In Bahariah et al.’s (2023) work, two CRISPR/Cas9 constructs carrying two gRNAs, targeting EgPAT or EgFAD2 gene, and one CRISPR/Cas9 construct consisting of all four gRNAs were successfully constructed. Meanwhile, depending on the position of seven gRNAs in the EgPDS genomic sequence, Jamaludin et al. (2023) successfully created two CRISPR/Cas9 constructs, assembling two gRNAs and one CRISPR/Cas9 construct with three gRNAs.

Alternatively, an endogenous tRNA-processing RNase, Csy4 endoribonucleases, and ribozymes were also engineered to control multiple gRNAs from a single transcript so-called polycistronic transcript system (Xie et al. 2015; Cermak et al. 2017). These multiple gRNAs were separated by the spacer sequences in the constructed plasmid. The functional gRNAs will be released due to the cleaving of the primary transcript by the respective enzymes (tRNA, Csy4, or ribozymes) at spacer sequences–gRNA junctions. Therefore, these polycistronic transcript systems can be utilized for multiplex genome editing in various living organisms, including plants (Ma et al. 2019; Huang et al. 2020; Li et al. 2021; Yang et al. 2022). Subsequently, it enables gene editing of multiple members of a gene family, genes with multiple functionalities involving a complex trait, or multiple sites of single genes. These polycistronic transcript systems are beneficial to editing oil palm genes involved in agronomical valuable traits usually regulated by multiple genes.

Transformation methods of CRISPR/Cas9 vector into plants

CRISPR/Cas9 genome editing in the plant required the delivery of DNA constructs consisting of Cas9 and gRNA into the cells. The delivery or transformation methods used in the recent applications of the CRISPR/Cas9 system in oil palm and other species are shown in Table 1. Yeap et al. (2021) reported the successful CRISPR/Cas9 gene editing in oil palm by targeting the EgBRI1 gene, which was determined using the biolistic transformation in immature embryos. Mutations were observed at the target sites of the EgBRI1 gene in oil palm shoots displaying dwarf phenotypes. Meanwhile, Bahariah et al. (2023) demonstrated that the biolistic transformation of embryogenic calli successfully produced oil palm plantlets with an edited EgPAT gene.

Other than that, Agrobacterium-mediated transformation can be another option for the delivery of CRISPR/Cas9 constructs into plant cells. This transformation method has been widely used for stable integration of CRISPR/Cas9 into plant genomes in many crop species, including apple (Charrier et al. 2019), banana (Zhang et al. 2022), potato (Ly et al. 2023), and plants listed in Table 1. Ali et al. (2015) and Yin et al. (2015) reported an improved Agrobacterium transformation utilizing the T-DNA embedded viral replicon, which was reported to enhance the NHEJ and HDR-mediated genome editing compared to the conventional T-DNA in tomato (Cermak et al. 2015). For oil palm, Bahariah et al. (2023) revealed a 28% mutation rate obtained through Agrobacterium-mediated transformation compared with a 6% mutation rate when embryogenic calli were transformed using biolistic. This finding suggested that Agrobacterium-mediated transformation is more efficient for altering the oil palm genome and, more importantly, for subsequently generating non-GMO gene-edited oil palms through crossing and segregation procedures.

Other than using protoplasts via polyethylene-glycol (PEG)-mediated transformation or callus via particle bombardment and Agrobacterium, immature embryos are currently being used for CRISPR/Cas9 studies. Additionally, immature embryos can be used for a DNA-free transformation system using RNP or mRNA (Svitashev et al. 2016; Liang et al. 2017). On the other hand, a new technology utilizing cell-penetrating peptides or CPPs has also been introduced. This strategy used infiltration of CPPs, positively charged short peptides that can move across cellular membranes of mature plant tissue. Interestingly, this technology has been shown to be capable of binding site-specific nucleases (Rádis-Baptista et al. 2017).

Recently, an efficient transformation method, namely a cut-dip-budding (CBD) delivery system, was developed to allow CRISPR/Cas9 genome editing in plants without going through the tissue culture process (Cao et al. 2022). This system delivered T-DNA of CRISPR/Cas9 gene construct from Agrobacterium strain into plant cells by infecting explants that can be generated to transformed shoots and subsequently to genome edited plants which was demonstrated in plants that recalcitrant to the genetic transformation such as dandelion Taraxacum kok-saghyz (Cao et al. 2023) and three succulent varieties Kalanchoe blossfeldiana, Crassula arborescens, and Sansevieria trifasciata (Lu et al. 2024) (Table 1). Hence, it would be intriguing to investigate a CBD delivery system for potentially editing the oil palm genome without requiring tissue culture.

Strategies of transient assay for developing CRISPR/Cas9 system

The use of transient assays to investigate the efficiency of the CRISPR/Cas9 system is highly recommended before utilizing the system in stable transformation (Shan et al. 2020). This is because stable transformation of recalcitrant plants such as oil palm is laborious and time-consuming compared to transient assays, which are more convenient and quicker. For now, two transient assay strategies are widely being used: protoplast transfection and leaf-cell agroinfiltration (Nekrasov et al. 2013; Xie and Yang 2013) (Fig. 3A). For example, Liang et al. (2017) reported using the transient system to test CRISPR/Cas9 efficiency in wheat protoplasts. Furthermore, for protoplast transfection, plant regeneration from the transfected protoplasts is also possible in some species, such as potatoes, tobacco, and lettuce (Andersson et al. 2018; Lin et al. 2018; Choi et al. 2022).

Fig. 3
figure 3

Strategies of transient assays and screening of mutation for CRISPR/Cas9 genome editing system. A Protoplast transient assay and agroinfiltration are commonly used for CRISPR/Cas9 proof-of-concept studies in new target species. B The CRISPR-induced mutation can be detected using SURVEYOR or T7EI assay, which uses enzymes sensitive to mismatched double-stranded DNA sequences. If indel is present, the heteroduplex containing the unpaired nucleotide (mismatch) will be cut by the SURVEYOR/T7EI enzyme, as shown in the gel image. C PCR/RE assay is based on the availability of the restriction enzyme in the CRISPR/Cas9 target site. The wild-type alleles will be digested in the presence of the restriction enzyme site. However, if the alleles are mutated, the restriction enzyme site is destroyed; therefore, the uncleaved DNA fragments will appear on the gel image, indicating the targeted mutagenesis by the CRISPR/Cas9 system

Full size image

In the protoplast transient assay, the steps include the isolation of protoplasts, transfection with CRISPR/Cas9 constructs, and protoplast culture. Generally, enzymes such as macerozyme and cellulase are used to remove plant cell walls to isolate protoplasts. The CRISPR/Cas9 constructs are delivered into protoplasts via various techniques, including electroporation, polyethylene-glycol (PEG), or microinjection (Lin et al. 2018; Wu et al. 2020; Zhou et al. 2020). Because of its simplicity and speed of the transformation process, the protoplast transient assay is an appealing model for evaluating the mutagenesis efficiency of a CRISPR/Cas9 system, particularly in assessing efficient gRNA. The appropriate expression of the CRISPR/Cas9 in the transfected protoplasts will potentially produce targeted editing in the sample (Andersson et al. 2018; Fan et al. 2020; Yu et al. 2021). For oil palm, the functionality of oil palm codon-optimized Cas9, and the effectiveness of the designed gRNAs were confirmed by an electroporation-mediated protoplast transient system (Yeap et al. 2021). The study revealed that oil palm codon-optimized Cas9 successfully promoted cleavage frequency of up to 25.49% in protoplasts transformed with gRNA targeting the EgPDS gene. Meanwhile, a significant result of 21% editing efficiency generating large DNA fragment deletion of 304 bp was obtained when oil palm protoplasts were transformed with CRISPR/Cas9 construct targeting EgFAD2 gene via PEG-mediated transformation (Bahariah et al. 2023).

Besides, protoplasts can be transformed with RNP comprising Cas9 and gRNA only (Svitashev et al. 2016; Badhan et al. 2021) (Table 2), which can avoid the integration of foreign DNA, resulting in non-transgenic plants. For example, using PEG-mediated delivery, Cas9/gRNA ribonucleoprotein was used to generate transgene-free lettuce (Woo et al. 2015) and potato (Andersson et al. 2018). On the other hand, the agroinfiltration method required the suspension of Agrobacterium containing T-DNA plasmid expressing the gRNA and Cas9 genes. Vacuum infiltration or direct injection is used to introduce the Agrobacterium culture into the plant’s tissues (usually leaves). The CRISPR/Cas9 components can be transferred through transfer DNAs (T-DNAs) into the host plant genome. Genome editing processes might then occur in the transformed cells. Numerous plant species, including Nicotiana, Solanum, and Lactuca, have established protocols for efficient and routine agroinfiltration (Wroblewski et al. 2005). Transformed cells can easily be detected, because the T-DNA segment contains a visual reporter gene such as green fluorescent protein (GFP), as Jiang et al. (2013) reported. Additionally, various researches have been published utilizing the Agrobacterium rhizogenes-mediated hairy root transformation for rapid analysis of CRISPR/Cas9 efficiency in many legume species (Belhaj et al. 2015; Barman et al. 2020).

Table 2 Summary of CRISPR-RNP applications in various plant species
Full size table

Screening of mutation and identification of mutant’s genotype

Several methods are used to detect the CRISPR/Cas9-induced mutation in various species, including oil palm (Table 1). These methods include polymerase chain reaction/restriction enzyme (PCR/RE) assays, T7 endonuclease I assay (T7EI), SURVEYOR nuclease assay, high-resolution melting (HRM) analysis-based assay, and PAGE-based genotyping assay (Shan et al. 2014; Thomas et al. 2014; Zhu et al. 2014). Although these assays are commonly used to screen for induced mutation, they have limitations, such as labour-intensive, time-consuming, sequence-limited, and expensive. The SURVEYOR assay and T7EI assay are commonly being used to detect mutation in CRISPR-treated samples, because these methods are more straightforward, cheaper, and applicable for any target sequence as both depend on the detection and digestion of mismatched heteroduplex DNA; however, their detection sensitivity is very low (Fig. 3B). In contrast, the PCR/RE assay is more reliable in mutation detection with a more straightforward methodology but requires restriction enzyme sites near the target sequences (Cong et al. 2013; Shan et al. 2014) (Fig. 3C). The method of PAGE-based genotyping assay is time-consuming and insensitive as it includes the denaturation and annealing of the PCR amplicons containing the induced mutations and detection using the native PAGE (Zhu et al. 2014). The HRM analysis-based technique is more efficient than PAGE-based assay as it depends on the melting temperature (Tm) distinction between the edited and non-edited PCR amplicons; however, it requires expensive equipment (Thomas et al. 2014).

Various software tools have been developed to detect and identify the CRISPR/Cas9 site-targeted mutation using Sanger sequencing data by analyzing the chromatogram files of Sanger sequenced-PCR products. The software tools are freely accessible and allow simultaneous analysis of multiple chromatograms in a short time. The most frequent software used are ICE (Interference of CRISPR Edits), TIDE (Tracking of indels by decomposition), TIDER (Tracking of Insertion, Deletions, and Recombination events), DsDecode (The Degenerate Sequence Decode), and CRISPR-Detector (Table 1). For example, ICE is the preferred software to quantify the presence of indels in the samples consisting of edited and unedited DNA mixtures, especially at low editing efficiency (Hasley et al. 2021). In addition, ICE analysis of Sanger sequencing data has shown a comparable or higher accuracy than next-generation sequencing (NGS). ICE used the untransformed or control sequences as the wild-type genomic reference to detect the presence of indels in the transformed samples. For oil palm, Bahariah et al. (2023) used ICE to evaluate the editing efficiency, indels type, and mutation genotype for samples derived from protoplasts transformed via PEG-mediated transformation, and embryogenic calli transformed via Agrobacterium- and biolistic-mediated transformations. The report emphasizes that ICE was a suitable software tool for the detection of mutations in gene-edited oil palm compared with other available software.

On the other hand, Yeap et al. (2021) employed TIDE for the detection of mutations in oil palm. TIDE is an assay that accurately determines the identity of indels and the frequency of induced mutagenesis in a cell pool by analyzing two resulting raw sequences obtained from one pair of standard PCR and Sanger sequencing reactions (Brinkman and Van Steensel 2019). TIDER is an upgraded version of TIDE that can predict the frequency of targeted mutagenesis, especially for small nucleotide changes induced by CRISPR homology-directed repair that uses a donor template (CRISPR-HDR) (Brinkman and Van Steensel 2019). TIDER software requires an additional sequencing trace, which can be prepared using a simple two-step PCR reaction. The DsDecode program can automatically decode the sequencing chromatograms with homozygous mutations, heterozygous or biallelic, into allelic sequences (Xie et al. 2019). This software can simultaneously read up to 30 chromatograms in a few minutes. For analyzing whole-genome sequence (WGS), which consists of both small and large datasets, CRISPR-Detector uses FASTQ sequencing files from control and CRISPR-treated samples (Huang et al. 2023). This software will process the data by aligning and sorting the reads, identifying indels, substitutions, and structural variants (SVs) while removing the genetic background and finalizing output to present the results.

The term mutation frequency is often used to describe the percentage of CRISPR/Cas9-induced mutation regenerated plants in which the targeted mutation can be detected at the locus or loci of interest (Shan et al. 2020). Previously, high mutation frequencies have been reported in rice and Arabidopsis (Ma et al. 2015). The genotype of the regenerated mutants depends on the time for the targeted CRISPR/Cas9-mutagenesis event to occur. Suppose a mutagenesis event happens before the first embryogenic cell divides. In that case, that is, in the early phase of plant regeneration and the locus on only one of the two sister chromatids was mutated, a diploid plant may be heterozygous or homozygous if both alleles were mutated with the same mutation or biallelic if both alleles were mutated. However, the breaks were repaired, resulting in different alleles (Shan et al. 2020).

However, in many cases, CRISPR/Cas9-targeted mutagenesis often occurs later in plant development and independently in different tissues (Jarvise et al. 2021). Thus, it produces chimeric plants with cells of different genotypes consisting of wild type, homozygous, heterozygous, or biallelic (Song et al. 2022). Early gene function studies can be conducted if the CRISPR/Cas9-induced homozygous and biallelic mutations occur in the first generation. This has been reported in tomato and rice (Brooks et al. 2014; Zhang et al. 2014). However, suppose the regenerated plants do not exhibit homozygous or biallelic mutations in the first generations or in the primary transformants. In that case, the phenotype analysis for loss-of-function activity can be analyzed in the later generation (Xu et al. 2015). This has been demonstrated in various species, including wheat (Wang et al. 2014) and Brassica napus (Yang et al. 2017), which obey the Mendelian heritability.

Proof of concept study of CRISPR/Cas9 system using PDS Gene

Establishing a CRISPR/Cas9 system in perennial plants such as oil palm can be more challenging than the annual plants. Many critical parameters, such as the ploidy level, genome heterozygosity, growth cycle, and physiological characteristics of the species, need to be carefully considered (De Bruyn et al. 2020). A high ploidy level will cause an increased workload for this gene-editing system to edit all copies of the target gene (Shan et al. 2020). Another challenge in introducing CRISPR/Cas9 study in oil palm is the long growth and complex physiology. These two factors are related, because long-growth cycle plants are usually perennials, primarily woody plants such as oil palms. Compared with small or annual plants, oil palms have a very long regeneration time; thus, the phenotyping of mutants can be complex. In addition, outcrossing and evaluating the inheritance of mutated alleles in the subsequent generations of most trees are tricky due to its dioecious nature (Bewg et al. 2018). Therefore, utilizing a target gene that can produce specific phenotypes at early plant development is advisable to identify the workable CRISPR/Cas9 system in oil palm (Bahariah et al. 2023; Jamaludin et al. 2023).

Phytoene desaturase (PDS) is the most utilized marker gene for CRISPR proof of principle studies (Shan et al. 2018). A marker gene is one in which mutants produce an obvious phenotype that allows easy visual screening. The PDS gene encodes an essential plant enzyme in the carotenoid biosynthetic pathway. It catalyzes the formation of one of the double bonds during the conversion of phytoene into lycopene. Therefore, additional copies of the PDS gene may increase the plant carotenoid content (Steinbrenner and Sandmann 2006). However, the most exciting criterion of the PDS gene is that it can be used as a photo-bleaching marker in plants, which can be easily exploited for proof-of-concept studies using CRISPR/Cas9. Silencing the PDS gene could disrupt the biosynthesis of carotenoid, chlorophyll, and gibberellins and produce albino plants (Jamaludin et al. 2023). Several studies have demonstrated the utilization of the PDS gene in various crops to test the efficiency of the CRISPR/Cas9 system (Fan et al. 2015; Yeap et al. 2021). In addition, PDS is typically a single copy gene in plant genomes, including oil palm, which is another advantage in its application.

Yeap et al. (2021) employed the EgPDS gene as a model gene to establish the CRISPR/Cas9 system in oil palm before applying the approach targeting the EgBRI1 gene. The group described the development of a transient protoplast assay to quickly define the efficiency of the CRISPR/Cas9 system in oil palm by targeting the EgPDS gene. From the five gRNAs tested, only two gRNAs, gPDS4 and gPDS5, successfully induced mutations in oil palm protoplasts with editing efficiencies of 83.33% and 62.50%, respectively. Using gPDS4 and gPDS5, they successfully generated oil palm shoots with chimeric albinism phenotype (Yeap et al. 2021). Unlike Yeap et al. (2021), who used single gRNA CRISPR/Cas9 constructs, Jamaludin et al. (2023) generated CRISPR/Cas9 constructs consisting of multiple gRNAs to modify the EgPDS gene in oil palm. However, no data from transformation works were reported in the study.

The model plant, N. tabacum, is an allotetraploid with a genome size of 4.5 G. It consists of high (> 70%) content of repetitive DNA (Sierro et al. 2014). It has become an essential model plant species for the study of CRISPR/Cas9 genome editing. Although N. tabacum contains four PDS genes (two each from the N. sylvestris and N. tomentosiformis progenitors), at which mutation of all four genes is necessary to obtain albino plants, various research has been successfully conducted utilizing tobacco as a model plant system for CRISPR/Cas9 study (Jiang et al. 2013; Nekrasov et al. 2013). The successful regeneration of the albino tobacco plantlet from the CRISPR/Cas9 transfected protoplast was also reported (Lin et al. 2018).

Oil Palm future prospect via CRISPR-RNP system

CRISPR/Cas9 is commonly implemented in plants to enhance agricultural qualities due to its ability to knock out genes. However, this procedure often involves transgenic intermediates that may raise legislation concerns and probably cause public rejection. Therefore, direct delivery of Cas9 protein and gRNA that can eliminate the integration of transgenes and reduce the risk of off-target activity, insertional mutagenesis, and immune response is more favorable (He and Zhao 2020). Previous research has shown successful delivery and CRISPR/Cas9 editing using laboratory-transcribed Cas9 and gRNAs into plant tissues (Woo et al. 2015; Klimek-Chodacka et al. 2021). This protein-based delivery system is also known as CRISPR–ribonucleoprotein (RNP) or CRISPR–RNP complex. This system requires two essential components in the formulation, the gRNA and the Cas9 protein, combined using a simple laboratory method to form a negatively charged recombinant Cas9 and gRNA or RNP complex. The RNP complex will then be delivered into the plant cell as one single reagent. In the cells, the RNP complex is susceptible to being degraded by both proteases and RNases, thus making it the most unstable format (Lin et al. 2022). However, for genome editing applications, the CRISPR/Cas9 reagents are only temporarily needed until the intended mutagenesis has happened; the presence of the CRISPR reagents is no longer favored as it can lead to off-target activity. Therefore, due to the rapid degradation of RNP in the cell, this format is more desirable to decrease the potential risk of off-target activity (Subburaj et al. 2022).

In contrast, when plasmid DNA is transformed into plant cells, the Cas9 gene may be randomly incorporated into the plant genome. The constitutive expression may increase off-target activity (Liang et al. 2018). These off-target effects can bring about the legislation concerns about GMO. The plasmid DNA or CRISPR/Cas9 expression vectors are usually delivered into the plant cells, where the Cas9 needs to be expressed first and then pre-assembled with the gRNA to form a complex, to finally able to cleave the target region, produce double-strand breaks and then, generate indels during the repair mechanisms. Unlike plasmid-based delivery system, RNP complexes only have to be pre-assembled by incubating the Cas9 protein with gRNA for 15 min at 37ºC in the laboratory (Klimek-Chodacka et al. 2021). The reaction produced a functional molecular complex that can be used directly for cellular genome editing. The RNP entered the nucleus using the NLS-fused Cas9, which is commonly used for RNP. Consequently, the action of RNP is more specific and quicker, since this complex can work immediately without needing intracellular transcription and translation in the genome (Liang et al. 2017). For these reasons, RNP is considered the most efficient CRISPR/Cas9 genome editing strategy for plants, including oil palm.

The delivery of RNP into plant tissues can be categorized into two approaches: physical approaches, such as microinjection, biolistic, and membrane deformation, or synthetic carrier approaches, such as lipid nanoparticles, inorganic nanoparticles, and nanogels (Zhang et al. 2021c). The PEG-mediated transformation has been the most commonly used technique to deliver RNP into the plant’s cells, including crop species such as apple and grapevine (Malnoy et al. 2016), rubber (Fan et al. 2020), banana (Wu et al. 2020), and recently in tomato (Lin et al. 2022), papaya (Elias et al. 2023), and oil palm (Norfaezah et al. 2024). PEG caused the protoplast to clump, thus promoting the proximal interaction between RNP and the cell surface (Masani et al. 2014). Previous research on protoplast PEG-mediated transformation has shown that an average of 10% editing rate was obtained using RNP (Svitashev et al. 2016; Andersson et al. 2018). Andersson et al. (2018) also reported that delivery of RNP using protoplast PEG-mediated transformation was efficient in editing all four copies of the PDS gene in 2%-3% of the regenerated shoots in potato species. Protoplast PEG-mediated transformation was also proven efficient in other crops, such as grapevine, apple, potatoes, lettuce, soybean, petunia, and wheat (Woo et al. 2015; Malnoy et al. 2016; Kim et al. 2018; Subburaj et al. 2022).

For oil palm, editing efficiencies of 63.6% up to 100% were obtained when protoplasts were transformed with multiple RNP, two combinations of RNP (RNP1/RNP2 or RNP3/RNP4), and three combinations of RNP (RNP5/RNP6/RNP7), targeting EgPDS gene (Norfaezah et al. 2024). The high editing efficiencies achieved in the study were due to the established efficient protocol for the preparation of RNP and protoplast transformation, including the two most influenced parameters, the amount of Cas9 protein and heat-treatment applied to transformed protoplasts. However, implementing protoplast PEG-mediated transformation in those plants is challenging due to the absence of regeneration protocols, and even if the protocols are available, the regeneration rate is meagre, such as in oil palm (Masani et al. 2013). Masani et al. (2014) also introduced the protoplast microinjection system for oil palm transformation work. However, no RNP delivery application through oil palm protoplast microinjection has yet been reported.

The Cas9 gene and gRNA can also be transformed directly into plant cells via particle bombardment. The RNP is coated with gold or tungsten microparticles before being bombarded into the target cell walls or membrane using helium shock (Stewart et al. 2018). This strategy produced high frequencies of mutated alleles in maize embryo cells, as Svitashev et al. (2016) reported. This strategy has also efficiently produced transgenic rice with inherited mutation (Shan et al. 2014). Optimization study of particle-bombardment delivery techniques into the plant cells using gold-activated silica nanoparticles has been described in detail by Martin-Ortigosa and Wang (2014). Hamada et al. (2017) also reported the utilization of in-planta particle bombardment (iPB) targeting mature plant tissue to avoid the regeneration of immature cells in wheat. This strategy is beneficial for those species that cannot be regenerated through the protoplast system but are amenable to callus regeneration. This approach was also successfully utilized in other species, such as wheat (Liang et al. 2017) and rice (Banakar et al. 2019).

Moreover, RNP can also be delivered via protoplast electroporation. This strategy exploited the disruption of cell membranes using an electrical pulse. Subsequently, producing temporary nanopores on the plant membranes will allow the transportation of RNP biomolecules (Boukany et al. 2011). Altogether, the DNA-free mutants produced using CRISPR-RNP technology are undifferentiated from the natural species. Therefore, they could be exempted from GMO legislation. Thus, CRISPR-RNP could be the new direction of futuristic breeding for sustainable oil palm agriculture. The summary of recent CRISPR-RNP applications in various plant species is shown in Table 2.

Conclusion and prospects

The advancement in CRISPR/Cas9 technology will enormously facilitate oil palm basic research, breeding, and functional studies of target genes. The development of CRISPR/Cas9 systems that are workable in oil palm would provide an easy-to-use platform to address diverse questions in oil palm functional genomics, such as the study of physiological traits, optimization of certain beneficial metabolic pathways, and genotype–phenotype relationship of oil palm genes. Site-specific mutagenesis by CRISPR/Cas9 system will enable deeper accessibility to gene expression study of various genes and targeting multiple loci despite the genomic complexity of the oil palm genome, which was previously difficult to access using the conventional molecular methods. Various platforms are being developed to increase the editing efficiency and specificity of the CRISPR/Cas9 system (Fig. 4), subsequently providing unprecedented potential for genetic studies of oil palm agronomical and economically beneficial traits. This includes higher proportions of desaturated fatty acids in palm oil, slow height increments, low lipase activity in mature fruits, improved nutritional values of palm oil, higher resistance to disease, and better environmental adaptability. The application of DNA-free genome editing, such as CRISPR-RNP, to produce DNA-free gene-edited oil palm may become an efficient strategy to address issues in current GMO regulations and public acceptance. This genome editing system will play a pivotal role in the sustainable agricultural development of oil palm.

Fig. 4
figure 4

Various approaches have been developed for establishing efficient and robust CRISPR/Cas9 gene editing system in oil palm (Masani et al. 2022; Fizree et al. 2023; Yeap et al. 2021; Bahariah et al. 2023; Jamaludin et al. 2023; Norfaezah et al. 2024). These approaches, (1) identification of suitable target genes and bioinformatics tools for designing efficient gRNA, (2) simple and efficient protocol to synthesise gRNA molecules and development of in-house robust CRISPR/Cas9 vector system, (3) development of multiplex genome editing system for editing valuable target traits such as oil palm palmitoyl-ACP thioesterase (EgPAT) and oleoyl-CoA desaturase (EgFAD2) genes for increasing oleic acid content, oil palm gibberellin acid 20 oxidase (EgGA20ox) genes for reducing tree height increment and oil palm virescens (EgVIR) genes for generating virescens fruits type, (4) efficient transformation, selection, and plant regeneration system for delivery plasmid-based CRISPR/Cas9 into embryogenic calli and protoplasts, (5) efficient RNP-based transformation and plant regeneration system utilizing oil palm embryogenic calli and protoplasts, and (6) evaluation of suitable screening method for detecting mutations in transformed plants, are in pipeline for developing a DNA-free genome editing protocol for oil palm improvements

Full size image