There are more than 4,000 landraces of native potatoes, mostly found in the Andean Region and over 180 wild potato species (source: International Potato Center). Despite the great number of potato varieties available, few cultivars are commercialized, chosen for their viability to be marketed and stored (Yang et al., 2015). Cultivars ‘Desiree’ and ‘Bintje’ were early favorites for transformation assays but nowadays, most potato cultivars are amenable to tissue culture and nearly all commercially important varieties can be successfully transformed with modifications of the standard protocols (Vinterhalter, 2008; Han et al., 2015; Halterman et al., 2016; Nadakuduti et al., 2018; Bruce and Shoup Rupp, 2019; Kaur et al., 2020). In 1988, it was reported the development of a genotype-independent method to transform four potato cultivars (‘Bintje’, ‘Berolina’, ‘Desiree’, and ‘Russet Burbank’) using leaf discs (De Block, 1988). However, subsequent evidence showed that the transformation efficiency is dependent on genotypes. There are several reports comparing the regeneration and transformation efficiencies of different genotypes using the same protocol and the results indicate that the effectiveness in obtaining transgenic plants is variable (Conner et al., 1992; Dale and Hampson, 1995; Kumar et al., 1995; Heeres et al., 2002; Han et al., 2015; Bakhsh, 2020). Moreover, Heeres et al. (2002) determined that regeneration and transformation efficiency are two different genetically controlled factors. So, for a given genotype, the success of transformation depends on several critical factors including vector, Agrobacterium strain, infection time, mode of injury, pre-culture period, cocultivation time, explant type, composition of the subculture media, selection markers, among others (Kaur and Devi, 2019; Bakhsh, 2020). Thus, standard protocols can be considered to transform a new genotype but then all parameters should be optimized to achieve adequate efficiency.

Once the use of a bacterial delivery method has been defined, the next parameter to determine is the strain to use since it affects the efficiency of the genetic transformation. In this sense, LBA4404 A. tumefaciens strain is the most used in potato transformation protocols (Stiekema et al., 1988; Tavazza et al., 1989; Trujillo et al., 2001; Heeres et al., 2002; Bakhsh, 2020; Mollika et al., 2020). However, alternative strains such as EHA105, GV2260, GV3101, and C58C1 have also been successfully used (Kumar et al., 1995; Banerjee et al., 2006; Han et al., 2015; Ahmed et al., 2018; Craze et al., 2018; Bruce and Shoup Rupp, 2019; Décima Oneto et al., 2020). Despite the existence of numerous reports, there is not a clear association between the strain and the genotype to be transformed; this means that for a particular genotype the most suitable strain must be experimentally tested.

Once the strain was selected there are three principal factors that affect transformation efficiency: the optical density of the bacterial suspension, the addition of a centrifugation step of the bacterial culture before inoculation and the length of the co-cultivation period. Considering the evidence reviewed, it seems that the optimal optical density values are between 0.5 and 0.8, depending on the construct and the growth medium. Some protocols indicate that centrifugation of the bacterial culture before inoculation may affect the viability of the bacteria, causing a drastic reduction of transformation efficiency (Beaujean et al., 1998; Banerjee et al., 2006; Décima Oneto et al., 2020). Co-cultivation period should be long enough to enable proper T-DNA transfer, but prolonged periods should be avoided to reduce tissue damage and somaclonal variation. Therefore, the optimal time is between 24 and 96 h, being 48 h the most reported co-cultivation period (Millam, 2006). Nonetheless, for each genotype is necessary to determine empirically not only the optimal strain but also the co-cultivation conditions to obtain the best transformation efficiency.

Another important choice is the vector to employ based on the transformation strategy selected. The type, the size, the regulatory elements, the selectable marker gene, the cloning efficiency, the cost and the availability, among other factors, must be considered.

The type and the size of the vector are directly related to the transformation strategy selected, mainly whether a shuttle (see Hellens et al., 2000) or a not binary vector is necessary. Regarding the regulatory elements, the constitutive and ubiquitous CaMV35S promoter is commonly used to express transgenes in potato; however, there are many reports of promoters successfully employed for specific objectives (Yevtushenko et al., 2004; Pino et al., 2007; Li et al., 2015; Nahirñak et al., 2019). Moreover, regarding a cisgenic approach, well characterized tissue-specific or inducible promoters have been reported from potato (Martini et al., 1993; Naumkina et al., 2007; Almasia et al., 2010). Another point to consider in the vector design is to avoid repeated regulatory sequences to prevent possible silencing effects.

The selectable markers generally used to identify transformed plant cells are genes encoding resistance to antibiotics or herbicides (Bruce and Shoup Rupp, 2019). The neomycin phosphotransferase II (nptII) gene, which confers resistance to the antibiotic kanamycin, is the most used in potato transformation protocols (Barrell et al., 2013). Barrell et al. (2002) evaluated the efficiency of different selectable markers using identical vectors for potato transformation. The effectiveness of recovery of transgenic lines was ranked as follows: kanamycin resistance > hygromycin resistance > phosphinothricin resistance > phleomycin resistance > methotrexate resistance (Barrell et al., 2002).

Other markers have been reported for potato transformation that do not involve the use of herbicides or antibiotics and use xylose or galactose as selective agents instead (Haldrup et al., 1998; Joersbo et al., 2003). These systems offer alternatives to the conventional ones, although their use requires further optimization. Visual markers genes such as glucuronidase, luciferase and green fluorescent protein have also been adopted, although to a lesser extent (Sidorov et al., 1999; Verhees et al., 2002; Rakosy-Tican et al., 2007). Alternatively, marker-free transformation of potato has also been reported. This approach is based on the recovery of transformed plants by PCR screening of plants regenerated without the use of a selection system (De Vetten et al., 2003; Ahmad et al., 2008). However, this approach resulted in the recovery of transgenic potato lines at low frequencies, with most of lines displaying insertions of undesirable vector backbone sequences and only a few lines containing the desirable single T-DNA insertion, which is another important criterion for commercialization of GE crops (Kondrák et al., 2006).

Concerning A. tumefaciens binary vectors, the arrangement of expression cassettes in the plasmid should also be contemplated. The selectable/screenable markers should be near the left border to facilitate the selection of events containing the entire cassette since the transference of the T-DNA is directed from right border to left border. Also, the inwards opposite orientation of the cassettes should be avoided preventing possible silencing effects.

Finally, it is important to check the vector through DNA sequencing before starting the transformation protocol.

For a genetic transformation system to be effective, it is essential to develop a rapid and efficient regeneration protocol. Leaves, stems, tubers, petiole, protoplasts, and micro-tubers have been used as explants to develop transgenic potato lines.

The source of explant tissue for potato transformation is frequently derived from in vitro plants (Barrell et al., 2013). The main advantage of using them is the supply of uniform and pathogen-free material for transformation (Newell et al., 1991). In vitro micropropagation of shoot cultures makes possible the accessibility of healthy and vigorously growing plant material throughout the year (Visser, 1991). Since it is sterile and already acclimatized to grow under in vitro conditions, surface sterilization of the plant tissue is not required, reducing both handling time and the possibility of contamination, and avoiding plant stress due to chemical treatments (Conner et al., 1992; Kumar, 1995).

Leaves and stem internodes from in vitro plants are the most widely used explants since they are readily available and easy to use. It has been reported that stem pieces are relatively robust and can be handled easily in larger numbers comparing to leaf explants, which are delicate and can be injured during the manipulation reducing the frequency of transformation and regeneration (De Block, 1988; Newell et al., 1991).

The main disadvantage of employing leaf and stem explants in transformation assays is the somaclonal variation which can occur in the callus phase (Visser, 1991). The use of tubers is advantageous because the possibility of somaclonal variation is reduced (Ishida et al., 1989). In vitro grown microtubers have several advantages over soil-grown tubers since they are derived from virus-free, aseptically grown potato shoots, they can be produced conveniently at any time in large quantities, and they take up less storage (Kumar, 1995).

It is important to note that a critical factor influencing the frequency of callus and regeneration is the physiological state of the starting material (Chakravarty and Wang-Pruski, 2010). To ensure a successful transformation, the selection of young and healthy explants from stock plant is preferred while damaged tissue will negatively affect regeneration potential (Ahmed et al., 2018; Craze et al., 2018). Moreover, the size of the explant should be large enough to resist Agrobacterium co-cultivation or biolistic bombardment without losing moisture during transformation procedures. Frequently the explant chosen for transformation depends on the cultivar or is determined by the experience of each laboratory (Barrell et al., 2013).

In conclusion, there are many options regarding tissue amenable for potato transformation, in any case it must be a healthy material, sterile, appropriate for manipulation and with regenerative capacity to obtain good and reproducible results.

Most protocols employ a two-step regeneration procedure, with a callus induction step followed by a shoot growth step. In particular, the de-differentiation and redifferentiation of explants is one of the main points of an effective plant transformation (Zhang et al., 2020). There are many types of plant tissue culture media, most based on Murashige and Skoog’s MS Media (Murashige and Skoog, 1962). MS media contains the major salts, a variety of minor salts, sucrose, vitamins, and plant growth regulators, which include cytokinins, auxins, gibberellins, abscisic acid, and ethylene (for details about regulators used in potato tissue culture see Bruce and Shoup Rupp, 2019). The callus induction stage is often facilitated by treatment of explants with zeatin or zeatin riboside with low levels of auxin, while the shoot induction stage often has a reduction of zeatin and auxin, plus the addition of gibberellin to stimulate shoot outgrowth. Regeneration rates per explant are usually high and the first shoots appear after 4–6 weeks (Millam, 2006), depending on the genotype used. For each potato genotype, it is necessary to adjust and determine the ratio of growth regulators most suitable for dedifferentiation and form callus, which also depends on the explant used (Zhang et al., 2020). Furthermore, another factor that influences the different transformation steps is light types, intensity and photoperiod (Gelaye, 2014).

On the other hand, the success of genetic transformation relies not only on the DNA delivery approach and a rapid and effective regeneration protocol, but also on the selection system to identify the transgenic cells. An efficient selection system results from a delicate balance between the inhibition of growth of wild-type cells and the preservation of the regeneration capacity of transformed cells. It is important to determine empirically the lowest level of the selective agent that prevent the development of non-transformed cells under selective conditions to reduce the recovery of false positives (also named “escapes” since they have escaped the selection process). Moreover, if the selection pressure is too high it may result in false negatives because of the loss of transformed plants (Barrell et al., 2002). Even though several selection strategies have been reported for potato transformation, kanamycin resistance is the most used selectable system and was shown to be more effective in the rapid recovery of large numbers of independently derived transgenic lines (Barrell et al., 2002). It has also a history of safe use in 122 transgenic crops approved for cultivation, food, or feed (ISAAA, 2021).

The freshness of the media and growth regulators is relevant so they must be replaced regularly. Moreover, experiments controls are very important since they allow to identify problems, for example, non-inoculated pieces of leaves placed with a selection agent to demonstrate their functionality, and without selection to ensure that the shoots can regenerate. The experience acquired to identify the putative transgenics trough visual selection is another main point in potato transformation protocols, for example, the formation of roots in the selective medium is an excellent indicator of the existence of transformed tissues.

As it was already mentioned, one of the main constrain in potato genetic engineering, is the somaclonal variation, which consists of phenotypic changes observed when plants are regenerated from cultured somatic cells. The observed phenotypic variations among somaclonal potato lines involve physiological, epigenetic, and/or genetic changes (Barrell et al., 2013). Maintaining potato lines in culture for prolonged periods can result in these variations (Craze et al., 2018). Genotype, explant origin and the culture conditions are others critical variables contributing to somaclonal variation (Meiyalaghan et al., 2011). The success of gene transfer techniques also depends on minimizing these variations (Stiekema et al., 1988), especially by decreasing the callus induction stage (Millam, 2006) or the period of in vitro culture in general (Han et al., 2015).

As it is shown in Table 1 even though there are many potato transformation protocols reported for different genotypes most of them are based on A. tumefaciens, employ nptII as selective marker, and use leaves and stems as tissue explant.

Like in all plant species, delivery of the genome editing reagents in potato is based on previously established transformation methods (Ran et al., 2017). Therefore, A. tumefaciens-mediated transformation has been the most widely used approach to deliver either TALEN or CRISPR/Cas systems in potato (Table 2). Genomic integration of transgenes encoding the genome-editing reagents is effective in producing a sustained expression of the components that led to the intended modification of the target site(s).

Transient expression of the genome editing reagents is a promising alternative delivery method to obtain transgene-free edited plants (Chen et al., 2019). This is of a special importance in a vegetatively propagated and highly heterozygous crop like potato since the elimination of the transgenes by crossing or self-crossing is not suitable. Transient expression of genome editing reagents has been achieved in potato using different strategies. Bánfalvi et al. (2020) have obtained edited potato plants of the tetraploid cultivar Desiree trough transient expression of a CRISPR/Cas9-coding vector upon A. tumefaciens-mediated infection by using a selection protocol consisting in 3 days-kanamycin treatments. Edited lines (2–10%) were screened to check the binary vector incorporation in their genomes by PCR amplification of specific vector fragments, with negative results in all analyzed cases (Bánfalvi et al., 2020). Nicolia et al. (2015) had developed a pipeline based on the transient expression of TALEN in potato by protoplast transfection with DNA vector followed by whole plant regeneration without selective agents, obtaining 10% of the regenerated lines with mutations in the expected region (Nicolia et al., 2015). In addition to the transient expression of DNA vectors, protoplasts allow the delivery of CRISPR/Cas9 components as pre-assembled ribonucleoprotein (RNP) complexes (Woo et al., 2015). This represents a promising alternative, with the potential of both reducing to a minimum (or even eliminating) the possibility of foreign DNA insertions in the plant genome and minimizing the probabilities of unwanted off-target effects. Such an alternative may result of great interest in the application of this technology to obtain potato commercial cultivars, since the obtained edited plants could be indistinguishable from those containing naturally or conventionally induced mutations (Kumlehn et al., 2018). The first application of RNP complexes to obtain potato edited lines was reported by Andersson et al. (2018). Authors transfected protoplasts isolated from the cultivar Kuras, with two different types of RNP complexes targeting the GBSSI gene and obtained a frequency of 1–25% of regenerated edited lines, varying with the transfection conditions and the origin of RNP complexes (Andersson et al., 2018). Furthermore, for one transfection up to 3% of the total analyzed lines displayed mutations in all four alleles of the target gene and no DNA insertions. Since this first report, RNPs have been successfully used to mediate potato genome editing using the protoplasts transfection and regeneration system, with editing efficiencies ranging from 27 to 68% (González et al., 2020) and 52 to 72% (Zhao et al., 2021), varying with the targeted genes and the design of RNP components. Protoplasts provide a suitable platform for genome-editing reagents delivery and may in some cases represent even a more efficient strategy when compared to Agrobacterium-mediated transformation (González et al., 2021). Even though it represents a promising strategy to mediate transgene-free genome editing in potato, the use of RNP in protoplasts present some important aspects to consider. For instance, DNA traces remaining in the assembled RNPs may led to unintended foreign DNA insertions in the regenerated plants, with different frequencies (Andersson et al., 2018; González et al., 2021; Zhao et al., 2021). A possible origin of such traces is the DNA molecules employed for the in vitro transcription of sgRNAs, before RNP assembly. To solve this issue, the use of synthetically produced sgRNAs to obtain the RNP complexes completely eliminates the foreign DNA insertions (Andersson et al., 2018). On the other hand, regeneration of complete potato plants from protoplasts has been associated with somaclonal variation and chromosome instability (Barrell et al., 2013; Fossi et al., 2019). Therefore, the selection of strategies that enhance the number of edited lines after protoplasts regeneration, ensures a vast number of individuals to select the best candidates for further phenotypic analysis (Andersson et al., 2017).

Clustered Regularly Interspaced Short Palindromic Repeats/Cas systems have been rapidly expanding and new CRISPR-related tools have been created that greatly broaden the scopes of genome editing and/or made it more precise.

Base editing technology is based on the fusion of a catalytically inactive (or partially inactive) Cas nuclease, fused to a cytosine or adenosine deaminase domain capable to convert one base to another (Mishra et al., 2020). Cytosine base editors (CBEs) convert a C to a G, while adenosine base editors (ABEs) mediate the conversion of A to G. CBEs have been successfully used to mediate potato base-editing (Table 2). Moreover, CBE consisting in a Cas9 nickase (nCas9) fused to either a human (Zong et al., 2018) or a Petromyzon marinus cytidine deaminase (Veillet et al., 2019a,b) were reported to obtain loss-of-functions or gain-of-function mutations related to important traits in potato.

The use of CRISPR/Cas systems originated from other bacteria species than S. pyogenes provide new benefits, including but not limited to, the expansion of the putative target sites within a given genome via different PAM requirements and the generation of different cutting patterns to introduce the DSB (Huang and Puchta, 2021). In a proof-of-concept publication, Veillet et al. (2020) demonstrated that Cas9 nuclease derived from Staphylococcus aureus (SaCas9) is efficient in inducing both genome editing via the NHEJ repair pathway and precise base editing in potato. Since SaCas9 recognize a more complex PAM (5′-NNGRRT-3′) it could additionally represent a highly specific nuclease by limiting the off-target activity, particularly for highly conserved genomic sequences in polyploidy species, such as potato (Veillet et al., 2020). This and other CRISPR/Cas orthologs systems are available for potato genome editing and can expand the toolbox for trait improvement (Huang and Puchta, 2021).

The recently developed prime editing technology allows the creation of different types of genomic changes with high precision, which represents potentially a new technological breakthrough (Gao, 2021). Prime editors are composed of an nCas9 fused to a reverse transcriptase, guided to the target site via a modified prime-editing guide RNA. In addition to define the target site, the prime-editing guide RNA serves as a template for reverse transcription, carrying a primer binding site and a sequence to be copied in the genome at its 3′end. Once nCas9 nicks the DNA, the released ssDNA can hybridize with the primer binding site and be used as a primer for the reverse transcriptase, which transfers the sequence encoded in the prime-editing guide RNA to the DNA strand (Huang and Puchta, 2021). The new sequence is incorporated later into the target site, through the DNA repair. Although prime editing is still inefficient in plant cells and no reports with this technology has been published in potato so far, its applications in a few crop species and its continuous optimization rise high hopes to incorporate it to the potato genome editing toolbox (Lin et al., 2021).

As we already stated in this review, genetic engineering includes genetic transformation and genome editing. Genetic transformation comprises the traditional tools for introduction of a gene of interest randomly integrated into plant genomes. Genome editing techniques have been developed as an alternative to introduce precise and predictable genome modifications into plants without adding foreign DNA. For a GE potato to become a market success it must be accepted by government regulatory system, producers, and consumers (Halterman et al., 2016).

Nowadays, most countries practicing commercial agriculture have established regulations, with different degrees of stringency, for field experimentation and later larger scale cultivation of GE crops. These regulations take into consideration food, feed, and environmental safety risks. Some countries have a process-oriented regulation and have established that the regulations that apply to genetically modified organism should be also applied to genome editing developments (European Union, New Zealand) (Nadakuduti et al., 2018). Other countries have pronounced in favor of regulating the varieties obtained by genome editing as conventional ones, if the developed varieties lack from any foreign genetic material (Argentina, United States, Brazil, etc.) (Feingold et al., 2018; Lema, 2019).

After more than 20 years of biosafety analysis of different GE potatoes, the main concern still regards the potential genetic flow between cultivated and wild relatives plants, particularly in the centers of origin. In most of the major potato producing countries, the chance of gene flow to interfertile species is virtually non-existent due to the lack of wild relatives. In South America, the Andean region is the center of origin of potato and represents a valuable source of diversity, being an important genetic resource. Even though intercrossings between cultivated potato and wild relatives have been reported, it is commonly restricted by reproductive barriers (Hawkes and Hjerting, 1969; Rabinowitz et al., 1990). Thus, the risk associated with a commercial released of GE potato in regions where interfertile species coexist should be evaluated case by case. In Argentina, an interspecific out-crossing trial between a S. tuberosum cv. Spunta transgenic line and the only wild relative species in the Pampean region that could possibly cross with it (Solanum chacoense), failed to detect any event of transgene transfer under natural field conditions (Bravo-Almonacid et al., 2012). In addition to safety, public concern also focuses on patents, plant breeders’ rights and the concentration of intellectual property in a small number of corporations (Haverkort et al., 2016). These are some points, together with the economic cost of risk assessment and regulatory framework process, that the political leadership should include in their schedule to facilitate the social access of these technologies to their farmers.

Regarding the producers, they are who probably see, more directly, the great benefits of the use of GE potatoes in their production. Recently, Bangladesh started a field trial of two late blight resistant transgenic potato lines, the success of which will allow 20% less yield loss and farmers will save around $ 12 million spent on fungicides (source PotatoPro³). In addition, the avoidance of agrochemical uses and its impact on carbon balance that may affect climate change will lead to important environmental benefits. Also, in Uganda, another transgenic variety that resists late blight infection without the use of fungicides is being evaluated. This would improve the safety of farmers and their families, and in particular the small farmers, who have limited access to fungicides, could reduce their production losses by up to 60% (Potato News Today, 2021). According to Ghislain et al. (2021) delaying adoption of this potato will lead to continued pesticide use and significant losses that will affect the most vulnerable farmers. On the other hand, the improvement of GE plant resistance to abiotic stresses such as salinity, drought, or temperature allows potato cultivation in less fertile agricultural lands (Hameed et al., 2018). Nevertheless, looking closely at the history of approved GE potato it can be noticed that producers may be reluctant to incorporate improved GE varieties because the industry fears mixing them with conventional ones and thus losing markets (Grafius and Douches, 2008; Haverkort et al., 2016; Tagliabue, 2018). This industry decision responds to a pre-judgment of consumers reactions, or to avoid negative publicity of their products. However, a science based potential risks and benefits communication campaign should be undertaken to comply with consumers’ right to be informed.

Consumer apprehension toward GE technology exists even though the cultivation area of these crops in 2020 was around 190 million hectares all over the world. According to Mullins et al. (2006), consumer concerns toward these technologies focuses on the public’s desire for unbiased information about potential risks. This leads us to think about the need to communicate clear, precise, and comprehensible information about GE technologies to the public, particularly showing the great advantages on its use. Public debate on the release of GE crops has led to questions regarding their environmental safety; however, it is important to clearly transmit that the deregulation of GE crops involves risk assessments and a complex regulatory framework. Concerning ecological compatibility of GE potato, many studies focused to address biosafety aspects and risk assessment did not find undesired ecological side-effects (Griffiths et al., 2000; Heuer et al., 2002; Rasche et al., 2006; Bravo-Almonacid et al., 2012; Fernandez Bidondo et al., 2019; van der Voet et al., 2019). In addition, as we mentioned before, GE potatoes resistant to biotic factors allow a radical reduction of pesticide applications, which are harmful to the environment and human health. Other advantage of GE potato directly relevant for consumers is the elimination of anti-nutritional or allergens to improve the nutrition quality. Recently, the potato variety Z6, engineered for low reducing sugars, low acrylamide potential, reduced black spot bruising and late blight protection, was deregulated in United States (source: USDA, 2021). It is interesting to discover if this GE potato with direct nutritional benefits will be well received by consumers. According to the international database, there are currently 51 events recorded on the potato list (ISAAA, 2021) and hundreds in the regulatory or pre-commercialization pipeline. The public concerns of consumers must be allayed with the information necessary to understand that GE potato are important for the sustainability of potato production, food security, income generation, and environmental protection. Therefore, it is worth taking a proactive approach to consumer apprehension in which all actors involved in the generation and adoption of these technologies provide that accurate information.

The Figure 1 summarizes the different strategies available for potato genetic engineering, including both genetic transformation and genome editing focusing on the common steps and the differences between them.