Author: Dídac Jiménez Sánchez Academic tutor: Irma Roig Vilanova
External tutor: Marta Pujol Abajo 15/ October/ 2021
Biosystems engineering Degree’s final project
Gene editing in melon (Cucumis
melo L.) using CRISPR/Cas9 to
functionally validate candidate
genes for ripening and fruit shape
Resum
La forma, qualitat i maduració del fruit són característiques importants en meló (Cucumis melo L.). Com que el meló posseeix una gran diversitat fenotípica, incloent tant varietats climatèriques i com no climatèriques, és un subjecte adequat per a l'estudi d'aquests trets. Amb l’objectiu d’identificar gens candidats de trets agronòmics importants en el meló, s'han iniciat diversos projectes per a localitzar quantitative trait loci (QTL) utilitzant poblacions de mapeig. Entre ells, s'han identificat dos QTL importants, FSQS8.1 i ETHQV8.1, per la forma i la maduració del fruit, respectivament. El gen candidat per FSQS8.1 és l’OVATE FACTOR PROTEIN (OFP1) i per a ETHQV8.1, l’ETHYLENE RESPONSIVE TRANSCRIPTION FACTOR (ERF024). CRISPR/Cas9 és un mètode punter d'edició genètica que podria ser una eina potent per ajudar a validar la funció de gens identificats. Amb aquesta finalitat en el laboratori s’havien generat DNAs recombinants que contenien gen Cas9 juntament amb un ARN guia específic per a cada gen. Aquests DNAs recombinants es van introduir en Agrobacterium tumefaciens i amb aquest es van transformar cotiledons de llavors de meló, dels quals es van regenerar plantes mitjançant cultiu in vitro. L'objectiu era obtenir plantes amb una mutació knock-out en aquests gens. Els fruits d'aquestes plantes s’autofecundaran per a obtenir mutacions en homozigosi i després es fenotiparan per a comprendre la funció dels gens candidats.
Durant aquest treball, es van obtenir amb èxit plantes editades amb CRISPR/Cas9. Algunes eren edicions bial·lèliques que estalviarien els esforços necessaris per obtenir un individu amb dos al·lels eliminats. No obstant això, el cultiu de teixits i la transformació amb Agrobacterium van presentar complicacions. Sis de les vuit transformacions no van tenir èxit. La principal problemàtica va ser la infecció d'una bacteri endogen. Les dues transformacions que van donar lloc a plantes editades només ho van fer després de d’entre 6 a 10 mesos de cultiu de calls. A més, totes les plantes obtingudes fins ara han sigut tetraploides, un problema comú en el cultiu de teixits de meló. Encara que CRISPR/Cas9 segueix sent una eina prometedora per a la investigació en meló, els mètodes de cultiu de teixits i transformació amb Agrobacterium s'han d’optimitzar i les dificultats del protocol d’edició CRISPR/cas s'han d'estudiar per resoldre-les.
Resumen
La forma, calidad y maduración del fruto son características importantes en melón (Cucumis melo L.). Ya que melón posee una gran diversidad fenotípica, incluyendo tanto variedades climatéricas como no climatéricas, es un sujeto adecuado para el estudio de estos rasgos. Para identificar genes candidatos para rasgos agronómicos importantes en el melón, se han iniciado varios proyectos para localizar quantitative trait loci (QTL) utilizando poblaciones de mapeo. Entre ellos, se han identificado dos QTL importantes, FSQS8.1 y ETHQV8.1, para la forma y la maduración del fruto, respectivamente. El gen candidato para FSQS8.1 es un OVATE FACTOR PROTEIN (OFP1) y para ETHQV8.1, un ETHYLENE RESPONSIVE TRANSCRIPTION FACTOR (ERF024). CRISPR/Cas9 es un método puntero de edición genética que podría ser una herramienta potente para ayudar a validar la función de genes identificados. Por lo tanto, para validar la función de los genes candidato, transferimos el gen Cas9 junto con un ARN guía específico para cada gen a través de Agrobacterium y regeneramos plantas mediante cultivo de tejidos. El objetivo era producir plantas con una mutación knock-out en dichos genes. Los frutos de estas plantas se autofecundarán para obtener mutaciones en homocigosis y luego se fenotiparán para comprender la función de los genes candidatos. Durante este trabajo, se produjeron con éxito transgénicos y plantas editadas con CRISPR/Cas9. Algunas eran ediciones bialélicas que ahorrarían los esfuerzos necesarios para obtener un individuo con ambos alelos eliminados. Sin embargo, el cultivo de tejidos y la transformación de Agrobacterium presentaron complicaciones. Seis de las ocho transformaciones no tuvieron éxito. Las causas de fallo fueron principalmente infecciones de una bacteria endógena. Las dos transformaciones que produjeron plantas solo lo hicieron después de 6 a 10 meses de cultivo de callos. Además, todas las plantas obtenidas hasta ahora eran tetraploides, un problema común en el cultivo de tejidos de melón.
Aunque CRISPR/Cas9 sigue siendo una herramienta prometedora para la investigación en melón, los métodos de cultivo de tejidos y transformación con Agrobacterium deben mejorarse y las dificultades del protocolo deben estudiarse.
Abstract
Fruit shape, quality and ripening are important traits in melon (Cucumis melo L.). As melon displays a vast diversity and even climacteric and non-climacteric varieties coexist, it is a suitable model for the study of these traits. To identify candidate genes for important agronomic traits in melon, several projects to locate quantitative trait loci (QTL) using mapping populations have been performed. Among them, two majors QTLs, FSQS8.1 and ETHQV8.1, have been identified for fruit shape and ripening respectively. The candidate gene for FSQS8.1 is an ovate protein factor (OFP1) and for ETHQV8.1 is an ethylene responsive transcription factor (ERF024). CRISPR/Cas9, a novel genetic edition tool, will help to validate the function of identified genes. Therefore, we transferred the Cas9 gene along with a guide RNA specific for each gene via Agrobacterium and regenerated plants through tissue culture. The objective was to produce plants with a knock out mutation in the target genes. The fruits of these plants will be selfed to obtain mutations in homozygosis, then phenotyped to understand the function of the candidate genes. During this work we successfully produced transgenics and CRISPR/Cas9 edited plants. Some were biallelic editions that would save the efforts needed to obtain an individual with both alleles knocked out. However, tissue culture and the Agrobacterium transformation proved to be challenging. Six out of eight transformations were unsuccessful. Failure causes were mostly due to infection from an endogenous bacteria. The two transformations that produced plants only did so after 6 to 10 months of calli culture. Further, all plants obtained so far were tetraploid, a common issue in melon tissue culture. Though CRISPR/Cas9 remains a promising tool for research in melon, tissue culture and Agrobacterium transformation methods need to be improved and the difficulties of the protocol need to be studied.
Table of contents
Introduction ... 7
Melon (Cucumis melo L.) ... 7
Ripening and genetic diversity ... 8
Fruit quality and shape ... 9
Breeding and genetic enhancing... 9
Plant tissue culture ... 10
Agrobacterium-Mediated Plant Transformation ... 11
CRISPR-Cas in plant biotechnology ... 12
Background of the project at CRAG ... 13
Objectives ... 16
Materials and Methods ... 17
Varieties and Introgression Lines used ... 17
Agrobacterium tumefaciens ... 17
Plasmids ... 17
Software ... 17
Kits ... 17
Culture media... 18
Buffers ... 20
PCR compounds ... 21
Melon transformation ... 22
Invitro regeneration and individualization... 24
Genomic DNA extraction ... 25
PCR ... 25
Agarose gel for DNA electrophoresis ... 26
PCR product purification (Sepharose) ... 26
RNA Extraction ... 27
Retrotranscription (RT-PCR) ... 28
Results and discussion ... 29
Approach ... 29
Obtention of OFP1 edited plants and sequence analysis ... 29
ERF024 transformation experiments ... 36
Conclusions ... 40
Bibliography ... 41
Table of figures
Figure 1: Showcase of the fruit diversity of Cucumis melo ... 7Figure 2: Range of shape and size in melon. ... 9
Figure 3: Scheme of CRISPR/Cas9 as a biotechnological tool. ... 12
Figure 4: Scheme of the CRISPR/Cas9 protocol to validate a plant gene in melon. ... 13
Figure 5: Scheme of the project to study the genomics of ripening and fruit quality. ... 14
Figure 6:Timeline of the project. ... 14
Figure 7: pDe-Cas9 plasmid. ... 18
Figure 8: Transformation protocol, day 3. ... 22
Figure 9: Transformation protocol, day 4. ... 23
Figure 10: Cuts applied to the seeds. ... 23
Figure 11: Transformation protocol, Friday. ... 24
Figure 12: Maintenance of the callus. ... 24
Figure 13: Final assembly of the tubes for the purification. ... 27
Figure 14: Forward sequence of an exon fragment of OFP1. ... 30
Figure 15: Individualized plants. ... 31
Figure 16: PCR electrophoresis results for Cas9 in plant samples. ... 31
Figure 17: Chromatograms from a section of OFP1 sequence... 33
Figure 18: Aligned sequences of OFP1 from all plants. ... 34
Figure 19: Contamination that appeared in the second attempt at transforming CALC8.1. ... 35
Figure 20: Electrophoresis results of the PCR for Cas9 in contamination colonies. ... 36
Figure 21: Forward sequence of an exon fragment of ERF024. ... 37
Figure 22: PCR electrophoresis results for Cas9 and ERF024 in calli samples. ... 37
Figure 23: Chromatograms of the callus sequencing from a section of ERF024. ... 38
Figure 24: Electrophoresis results of the ERF024 PCR test of the cDNA... 39
Table of Tables
Table 1: Germination medium. ... 18Table 2: Co-culture medium ... 19
Table 3: Selection medium ... 19
Table 4: Rooting medium ... 20
Table 5: CTAB buffer ... 20
Table 6: TAE buffer ... 20
Table 7: PCR Compounds. ... 21
Table 8: Primers used for PCRs and sequencing. ... 21
Table 9: Details of the sequencing results. ... 34
Acronyms
ACC 1-aminocyclopropane-1-carboxylate
ACO 1-aminocyclopropane-1-carboxylate oxidase ACS (Antibiotic) Acetosyringone
ACS (enzyme) 1-aminocyclopropane-1-carboxylate synthase
BA Benzylaminopurine
BP Base pair
CALC8.1 Piel de sapo introgression line carrying calcuta fragment in chromososme 8 Cas9 CRISPR associated protein 9
cDNA Coding deoxyribonucleic acid
CRISPR Clustered regularly interspaced short palindromic repeats CTAB Solution containing cetrimonium bromide
dNTP Deoxynucleoside triphosphate
ERF Ethylene responsive transcription factor
ERF024 Ethylene responsive transcription factor gene in melon
ETHQV8.1 QTL containing the ethylene responsive transcription factor gene in melon FSQS8.1 QTL containing the ovate family protein gene in melon
GMO Genetically modified organism gRNA Guide ribonucleic acid
IAA Indole-3-acetic acid
IL Introgression line
Indel % Percentage of sequences with a deletion KO-Score Percentege of deletions that cause a knock-out
MB Basal medium
MS Murashige and skoog
LB Liquid basal medium
OFP Ovate family protein
OFP1 Ovate family protein gene in melon
PCR Polimerase chain reaction
PPM Plant preservative mixture
PPT Phosphinothricin
PS Piel de sapo melon variety
QTL Quantitative trait loci
RIL Recombinant inbred line
TAE Tris acetate EDTA
TALEN Transcription activator-like effector nucleases
Ved Vedrantais melon variety
ZFN Zinc finger nuclease
Introduction
Melon (Cucumis melo L.)
Melon (Cucumis melo L.) is a very important fruit crop, ranking 12th in production in 2019 with 27 million tons produced that year (Food and Agriculture Organisation (FAO), 2020). Evidence of its cultivation exists as early as 3000 BCE in China, where it has been an important crop throughout its history (Walters, 1989).
It was also present in the first century CE in the Mediterranean (Janick et al., 2007). The Cucumis genus belongs to the Cucurbitaceae family, it comprises about 50 species including the cucumber (Cucumis sativus), also a very important crop. C. melo has a chromosome number of 2n=24 and a genome of 450 Mbp, which was sequenced and assembled in 2012 (available at http://melonomics.net) (Garcia-Mas et al., 2012). There are as much as 44.298 accessions of C. melo across many institutions of germplasm conservation (The Second Report on the State of the World’s Animal Genetic Resources for Food and Agriculture 2019). A large phenotypic variation exists within the species (Figure 1). The varieties are mostly classificated morphologically by flower and fruit. Cultivated melon fruits can grow to a variable range of weight (from 30 g to 35 kg) and shape, such as the round Cantalupensis group or the very long Flexuosus group, while wild melons are much smaller and oval. The exocarp of cultivars is extremely variable, through the combination of colors (e.g., white, green, yellow), vein tracts, ribs, blossom scars and textures (e.g., smooth, wrinkled, netted). Wild melons are usually green and smooth. Similarly, the mesocarp in cultivars ranges mostly between orange and white, and varies in sugar texture and aroma. Wild varieties have a light green flesh and are not as sugary as their cultivated counterparts. Variability also exists in the length and process of ripening of fruits, and so the shelf-life also varies (Pitrat, 2016).
Figure 1: Showcase of the fruit diversity of Cucumis melo. From (Pitrat, 2007)
Pangalo 1929 proposed a subspecies level intraspecific classificationClick or tap here to enter text.. It roughly separated eastern exotic varieties into the subspecies agrestis and cultivars into the subspecies melo. The classification was based on the pubescence of the ovary. Munger and Robinson, 1991 proposed classification through horticultural groupsClick or tap here to enter text.. Accessions are grouped using
fruit and flower character combinations of horticultural significance. This original classification has been revised and modified several times, proposing, merging or dividing groups. Pitrat 2016 proposed a subgroup levelClick or tap here to enter text.. A genome-wide single nucleotide polymorphism analysis revealed that, genetically, the subspecies are justified but divisions between groups were inconsistent and could not be considered botanical taxa though they are still useful (Jung et al., 2020).
Among all this variation, many traits are desirable for new lines and breeding programs. Some are fruit setting easiness, production rates, earliness, fruit quality, shelf life or pest resistance. Much research exists about breeding of varieties with enhanced traits through many methods including hybrid lines, Quantitative Trait Loci (QTL), breeding or genetic engineering (Kesh and Kaushik, 2021).
Ripening and genetic diversity
In plants that evolved to produce fleshy fruits their function is to disperse the seed through fruitarian animals. Seeds, however, are not necessarily ready for propagation at the early stages of fruit growth and so fruits have to shift from being unattractive and protective to being appealing to the vectors once the seed is ready. This process is known as ripening. It involves several physiological and biochemical changes such as starch degradation to accumulate sugars and organic acids, chlorophyll degradation along with pigment synthesis, production of aroma and tissue softening (Klee and Giovannoni, 2011). Perceived fruit quality by the consumer, especially flesh firmness, taste and aroma, varies depending on the stage and type of ripening. In addition, not all varieties ripe at the same rate, which affects shelf life (Farcuh et al., 2020).
Fruits are classified depending on the role the hormone ethylene has in ripening. Ethylene is a plant hormone that mainly regulates leaf abscission, senescence and ripening (Iqbal et al., 2017). Species or varieties are considered climacteric if the fruit shows a burst in respiration coordinated with a peak in ethylene at the onset of ripening, or non-climacteric, when they do not show such behavior (Biale, 1964).
Climacterism is however a complex quantitative trait and as such there is a wide range in the ripening processes and variation between and within species.
Ripening is a complex process that requires to be closely coordinated with seed development and it is regulated in the fruit through hormones, transcription factors and epigenetic mechanisms. Previous ripening studies have focused on tomato (Solanum lycopersicum) due to its inbreeding tolerance, efficient greenhouse propagation, short life cycle, ease of transformation and available full genomic sequence.
Several genes encoding the enzymes 1-aminocyclopropane-1-carboxylate (ACC) synthase (ACS) and ACC- oxidase (ACO) of the ethylene biosynthesis pathway were differentially expressing during fruit ripening, while also being regulated by ethylene. Ethylene independent pathways also play a role in ripening (Pech et al., 2012). This leads to a complex two-stage system where ethylene first represses and later promotes ripening, leading to the ethylene peak (Honda et al., 1972). DNA methylation, an epigenetic control factor also has an effect on ripening, where inducing whole genome hypomethylation provokes early ripening onset (Lang et al., 2017).
Melon is of interest to understand the ripening process as many varieties exist across the spectrum of climacteric ripening. For example, the C. melo var. cantalupensis ‘Védrantais’ (Ved) is fully climacteric while the C. melo var. inodorus ‘Piel de sapo’ (PS) is non-climacteric.
Fruit quality and shape
Fruit shape is considered a quality trait, since consumers can judge it a priori to choose from the available varieties. Shape refers to many parameters, mainly diameter and length and great variance exists between the accessions, using two extremes as example, fruits from the agrestis subspecies are oval and small in contrast with the flexuosus group, which are large and long (Figure 2).
Figure 2: Range of shape and size in melon. Left: round fruits from the subsp. agrestis. Right: Elongated fruit from the group
"flexuosus" and subgroup "tara". From (Pitrat, 2016).
Shape is also a polygenic trait, involving several gene families. These are: SUN, OFP, YABBY, CNR, KLUH/CYP78A and WOX (Monforte et al., 2014). In other species, Ovate Family Proteins (OFP) in conjunction with other proteins regulate the growth of fruits to a correct shape and also the growth in response to external stimuli or stress, and mutations on these genes produce pear shaped or flat fruits (Liu et al., 2002; Snouffer et al., 2020). Many quantitative trait loci (QTL) have been found for melon fruit shape (Kesh and Kaushik, 2021).
Breeding and genetic enhancing
Plant breeding is a technology focused on improving plant varieties by making directed changes to traits or adding new ones. For example, plants are bred to increase yield, quality, to resist pests, herbicides and other stress conditions or even to produce new marketable varieties. This is possible through a diverse collection of methods. Conventional breeding uses natural variability and normal reproductive methods.
From the available germplasm, it selects with different criteria the individuals with the desired traits and generates different populations. This includes breeding by selection, obtaining variance by artificial mutations, generating hybrid lines and many crossing strategies such as single-seed descent or
backcrosses. Molecular breeding has improved conventional breeding with biotechnology, by using molecular markers to help the introgression of desired traits. Biotechnology has also an important role to generate variability, targeting genes directly or introducing transgenes, among others (Baenziger et al., 2006).
To study this variability and identify genes responsible for traits, often, QTLs must first be identified. A QTL is a part of the genome whose variations correlate with a trait. To identify a QTL, mapping populations are used. These are populations originate from the cross of two different varieties that are then backcrossed through specific strategies. Two relevant types of mapping populations are inbred lines (ILs) and recombinant inbred lines (RILs). An IL is a population where each individual has the same combination of alleles from the two parentals, with every locus in homozygosis. A RIL is a population where each individual has a different combination of alleles from the two parentals, with every locus in homozygosis.
Each plant has then a combination of alleles from the two varieties. Then through molecular markers the combination of each plant is identified. The plants of the population are also phenotyped. With this data and through a complex statistical analysis, zones of the genome that explain large variations of a trait depending on which allele is present. Said zone is then identified as a QTL for the trait. Then, the genes presence in the sequence are the candidates to be underlying the QTL and thus have a function related to the trait (Abiola et al., 2003).
Plant tissue culture
Plant tissue culture (also called in vitro culture) is the science and technology used to grow plant cells, tissues, organs or individuals from explants. Explants are tissue fragments extracted from a “mother” plant and are grown on artificial media, neither on soil nor within the original plant. In vitro growth can be organized, when the integrity and structure of the original tissue is maintained, when organs are developed de-novo (a process called organogenesis) or when a somatic embryo is developed (non-zygotic, the process is called embryogenesis). In vitro growth can also be non-organized when the plant cells remain dedifferentiated, usually happening when the original explants cells were not pluripotent. Many types of culture exist such as shoot, meristems, protoplast or cell suspension culture. A common type is callus culture. A callus is an aggregation of dedifferentiated cells that grow and multiply without a structure, though not necessarily homogeneously. Any explant can form a callus with the correct growth regulators in the medium. From these formations, cells can differentiate again and regenerate into a normal plant or produce a somatic embryo. The mediums used are a concoction of: (1) macronutrients (nitrogen, potassium, calcium, phosphorous, magnesium and sulphur); (2) micronutrients (iron, nickel, chlorine, manganese, zinc, boron, copper and molybdenum) in lesser concentrations and not necessarily all used; (3) sugar, normally sucrose; (4) vitamins, especially thiamine; and (5) growth regulators, substances that regulate instead of nurturing growth. Depending on many factors and the purpose of the culture different concentrations and proportions are used. Auxins such as indole-3-acetic acid (IAA) regulate cell division and elongation and are involved in the formation of meristems, in the polarity or tropisms of the plant and also in inducing apical dominance. Cytokinins such as 6-benzylaminopurine (BA) regulate protein synthesis and the cell cycle and promote, instead of apical dominance, growth of lateral buds. Many other types of hormones exist and can be used for a myriad of purposes. The most used medium formula is the one developed by Murashige and Skoog in 1962 often called “MS” (George et al.,
2008). A component common to virtually all in vitro techniques is disinfection of explants and maintenance of sterility. The explant will invariably have microorganisms on themselves which usually find the types of mediums used very favorable and will compete with the explant and therefore thwart growth. To prevent this, explants are disinfected with chemicals to eliminate “exogenous” microbes (found on the surface, as opposed to “endogenous” microbes), the medium and tools that will be in contact with the explant are kept in sterility and antibiotics and fungicides are added to the medium.
In melon many tissue culture protocols have been developed. Both organogenesis and embryogenesis has been achieved from cotyledons, hypocotyls, roots, leaves, protoplasts, and shoot meristems. Cotyledons and other seedling parts are the most efficient for both organogenesis and embryogenesis (Debeaujon and Branchard, 1993; V. Moreno et al., 1985). Due to the great variability present in melon, genotype is the primary factor affecting regeneration successfulness. Different cultivars can have different responses to the same protocols (Molina and Nuez, 1995). Several challenges remain across in vitro culture in melon.
Somaclonal variations are common in melon plants originating from tissue culture and are often not desired (Ezura et al., 1992). Hyperhydration, as in other species, is a common issue present in melon when cultured in-vitro. It is a series of abnormalities often characterized by a lower chlorophyll content, high water content and irregular stomata among others. Hyperhydrated plantlets are usually no longer viable (Nuñez-Palenius et al., 2008).
Agrobacterium-Mediated Plant Transformation
Agrobacterium is a genus of bacteria responsible for the crown gall disease and capable of transferring segments of a Ti-Ri plasmid of about 200 kb into plant genomes using carrier proteins coded by vir genes.
In nature the bacteria transfers oncogenes, which activate the synthesis of hormones responsible for tumorous growth, and genes responsible of opine synthesis, which the bacteria uses as a carbon and nitrogen source. The species Agrobacterium tumefaciens is the most commonly used as a biotechnological tool. In practice, a binary vector system is used. In this system a “vir helper” plasmid containing the vir genes but no transfer DNA is used in combination with a “binary vector” with the desired genes, regulatory sequences and selectable markers. Agrobacterium “disarmed” strains, with only the vir helper plasmids such as Agl-0 are available for researchers (Lazo et al., 1991). These strains are transformed themselves and then allowed to infect the plant cells (Baloglu et al., 2018). After infection with the bacteria, in vitro tissue culture is required, which supposes a problem in many plants (Al-Khayri et al., 2016).
Agrobacterium mediated transformation protocols are available for C. melo. Transformation efficiency averages are reported to be between 2% and 7% (Dong et al., 1991; Fang and Grumet, 1990; Nonaka &
Ezura, 2015). Many traits have been genetically engineered in melon. Among the traits is disease resistance. Resistance to cucumber mosaic virus as well as zucchini mosaic virus was achieved by transferring a coat protein of the virus (Fang and Grumet, 1993; YOSHIOKA et al., 1993). Quality traits have also been experimented with, for example, the antisense ACO transformed melon plant delayed or removed the effects of climacteric ripening in cantaloupe melons (Ayub et al., 1996).
CRISPR-Cas in plant biotechnology
Previous to the discovery of CRISPR/Cas9, the available genome edition tools were the zinc finger nucleases (ZFN) and the transcription activator-like effector nucleases (TALEN). ZFNs are artificial dimer enzymes comprised of a zinc finger domain able to bind specifically to a sequence and a nuclease. TALENs are similar, but the binding domain is of different nature. However, the zinc fingers domain is considered to be complex to design. TALENs are easier to design but are susceptible to DNA methylation and are not as efficient as CRISPR/Cas9. The CRISPR/Cas9 system comes from a bacterial defense mechanism against virus. These bacteria, after a survived infection, place the foreign DNA (20 bp) into the CRISPR (clustered regularly interspaced short palindromic repeats) region. In the event of future invasions, the transcript of said DNA (crRNA) together with “trans activating RNA” (tracrRNA) forms a complex with a Cas protein, which is able to make double stranded breaks. The crRNA then hybridizes with the foreign DNA and is cut by the Cas at a specific location, close to a protospacer adjacent motif (PAM sequence). This mechanism is used to cleave any DNA given it contains a PAM sequence by providing a complementary sequence as crRNA. The PAM sequence for the Cas9 protein is “NGG”, where N is any nucleotide. Moreover, by combining the crRNA and the tracrRNA into a “guide RNA” (gRNA) only the Cas9 gene and the gRNA sequence with a promoter is needed to disrupt a gene on a host cell (Figure 3) (Jinek et al., 2012). CRISPR is also feasible in plants. To disrupt targeted genes in plants a plasmid containing both the Cas9 gene and the gRNA sequence is transferred using Agrobacterium. (Jiang et al., 2013).
Figure 3: Scheme of CRISPR/Cas9 as a biotechnological tool.
Hoogsvort et al. 2019 demonstrated its applicability in melon and set up a protocol. The phytoene desaturase gene was targeted and mutant plants were albino and dwarf. An 8 % of the explants (cotyledons) regenerated into shoots and of those 71 % were transgenic, though none had a biallelic edition This is our method to validate genes (Figure 4) (Hooghvorst et al., 2019).
.
Figure 4: Scheme of the CRISPR/Cas9 protocol to validate a plant gene in melon.
Click or tap here to enter text.The European Union legislation regarding Genetically Modified Organisms (GMOs), defined them as "organisms in which the genetic material (DNA) has been altered in a way that does not occur naturally by mating or natural recombination". This legislation is very strict and involves an expensive a lengthy process for authorization of the commercialization of new varieties. This definition includes transgenics and CRISPR mutants but not varieties obtained through mutagenesis or marker assisted breeding (Plan & van den Eede, 2010). In consequence biotech crop production in Europe is very low, of only 0.13 million hectares compared with USA’s 75 million hectares (Wiedenheft et al., 2012).
Background of the project at CRAG
I developed this final degree’s project the Centre for Research in Agricultural Genomics (CRAG) located at Bellaterra, Barcelona, in the group of Genetics and genomics of vegetable crops. This group studies ripening and fruit quality in melon through genomic and genetic tools. With such objective the group developed mapping populations to identify QTLs, sequenced the genome of melon and developed a CRISPR/Cas9 genome edition protocol for melon (Garcia-Mas et al., 2012; Hooghvorst et al., 2019). The group used RIL mapping populations from a PS × Ved (Pereira et al., 2018), introgression lines (ILs) from the cross PS × PI 161375 (“Songwhan Charmi”) (E. Moreno et al., 2008a) and ILs from the cross PS x Ved (Pereira et al., 2020). It is also an objective of the group to identify and validate the underlying gene of the most relevant QTL of each analyisis. This project is summarized in Figure 5 and a timeline is provided in Figure 6
Figure 5: Scheme of the project to study the genomics of ripening and fruit quality.
Figure 6:Timeline of the project.
Eduardo et al., 2005 developed a collection of IL using double-haploids from a PS × PI 161375 (“Songwhan Charmi”) cross. An IL named “SC3-5b” presented climacteric behavior. However, both parentals, are non- climacteric. Moreno et al., 2008 and Vegas et al., 2013 identified two QTLs: ETHQB3.5 and ETHQV6.3.
These QTLs reportedly interact with each other, with ETHQV6.3 having a greater effect. A further mapping population from SC3-5-1 (also carrying ETHQV6.3 and ETHQB3.5) × PS allowed to identify MELO3C016540 (CmNAC-NOR) as underlying gene candidate. Ríos et al., 2017 demonstrated that CmNAC-NOR was the causal gene when ‘Charentais mono’ TILLING mutants for CmNAC-NOR showed a significant delay at the onset of the climacteric ripening when compared to wild type.
The group also developed a collection of RILs from a Ved × PS cross started in 2008. Pereira et al., 2018 identified 33 QTLs related to ripening and fruit quality. The traits controlled by the QTLs were sugar and carotenoid content and morphology of fruit and seed. Pereira et al., 2020 performed another QTL analysis from the same RIL this time revealing 74 QTLs for ripening associated traits. The studied traits were ethylene, aroma and chlorophyll parameters, among others. ETHQV8.1 was identified as the most prominent one. Two introgression lines were developed: one containing the Ved ETHQV8.1 allele in a PS background and another containing the PS ETHQV8.1 allele in a Ved background. The Ved allele was sufficient to trigger climacteric ripening and in the opposite case, the PS allele delayed and reduced the ethylene production. Afterwards, ETHQV8.1 was fine mapped and of 14 candidate genes, 3 were
promising: (1) A serine/threonine kinase, CTR1-like (MELO3C024518) that inhibits ripening pathways unless ethylene is present; (2) A demethylase, ROS1 (MELO3C024516), whose orthologue in tomato has shown to affect ripening; (3) An ethylene-responsive transcription factor, ERF024 (MELO3C024520) . ERF proteins are usually involved in stress responses. In grape (Vitis vinifera) ERF genes were described to be upregulated in fruits that showed a faster ripening after they were exposed to higher temperatures (Carbonell-Bejerano et al., 2013; Thirugnanasambantham et al., 2015).
In this project we also worked al CRAG in collaboration with the genomics in plant breeding group of the Institute for Plant Molecular and Cell Biology in Valencia. Diaz et al. 2014 used an IL from a PS x PI124112 (mormodica group) cross started in 2002 as a mapping population. Ten QTLs were found: five for fruit length, two for fruit diameter and three for fruit shape. Regarding the fruit shape QTLs, FSQS8.1 was the most significant one. The PI124112 allele produces rounder fruits though with a significant epistatic interaction with the sex type gene ‘a' (‘A’) (Díaz et al., 2014). FSQS8.1 was fine mapped and an OFP candidate gene, CmOFP1, was found. This gene has a homologue in tomato involved in fruit shape (Wu et al., 2018).
Objectives
The main objective of this Bachelor’s degree final project is to validate the putative function of two genes in melon (Cucumis melo L.) by CRISPR/Cas9. These genes are: transcription repressor Ovate Family Protein 1 (OFP1), related to fruit shape, and Ethylene Responsive Transcription Factor 024 (ERF024), related to fruit ripening. To achieve this main objective, the specific objectives are:
-Transformation of melon seeds with Agrobacterium tumefaciens containing recombinant DNA -In vitro regeneration of transformed calli to individualized plants
-Analysis of individualized plants to check the alleles of the studied genes
Prior to the work described in this project, coworkers of the group cloned the recombinant DNA into the Agl-0 strain of Agrobacterium. Additionally, this type of experiment requires waiting for the plants to grow and produce fruit Acclimatization. Therefore, selfing and phenotyping are not part of the objectives of this project due to the time constraints.
Materials and Methods
Varieties and Introgression Lines used
For both experiments, we used seeds obtained during 2020 of the pure line Vedrantais (Ved) were used.
Vedrantais is round and climacteric.
For the fruit shape experiment, apart from transforming Ved seed, I used an Introgression Line (IL) carrying a fragment of PI 124112 (CALC) in the Piel de Sapo (PS) genetic background. This line was called CALC8.1, as the introgression from CALC was located in chromosome 8, containing the CALC allele of OFP1 gene. It produced round fruits. It was provided by the Institute for Plant Molecular and Cell Biology in Valencia (Díaz et al., 2014).
Agrobacterium tumefaciens
The Agrobacterium tumefaciens strain used in the transformation was the Agl-0.
Plasmids
The recombinant DNA of the plasmids transferred to the Agrobacterium to perform the transformation is detailed in Figure 7.
Software
To analyze the files provided by the sequencing service, we used “CodonCode Aligner 9.0” by CodonCode Corporation and “Synthego ICE analysis” by Synthego.
Kits
To extract RNA we used a Spectrum® Plant Total RNA Kit by Sigma Aldrich and to obtain cDNA we used a PrimeScript® Takara kit by Takara.
Figure 7: pDe-Cas9 plasmid. In color, the genes of the plasmid. In bold and outside, the restriction enzyme sites. It was transferred to the Agrobacterium once the guide sequence was placed between the attR loci.
Culture media
We fixed the pH of all culture media to 5.8 before adding the agar and we autoclaved it at 120 °C for 20 min before we added the hormones and antibiotics/biocides. The composition of all media used is detailed in tables 1 through 4.
Table 1: Germination medium.
Germination Medium
Compound Final concentration
Duchefa M0222 MS +Vit 4,4 g/l
Sucrose 30 g/l
Agar 8 g/l
Table 2: Co-culture medium
Co-culture medium
Compound Stock concentration Final concentration
Duchefa M0231 MS + B5 4,4 g/l
CuSO4 5H2O 1 g/l 1 mg/l
MES 0.6 g
Sucrose 30 g
Agar 8 g
6-Benzylaminopurine (BA) 1 g/l 0.5 mg/l
Indole-3-acetic acid (IAA) 2 g/l 0.1 mg/l
Acetosyringone (ACS) 400 mM 0.2 mM
Table 3: Selection medium
Selection Medium
Compound Stock concentration Final concentration
Duchefa M0231 MS + B5 4,4 g
CuSO4 5H2O 1 g/l 1 mg/l
MES 0.6 g
Sucrose 30 g
Agar 8 g
6-Benzylaminopurine (BA) 1 g/l 0.5 mg/l
Indole-3-acetic acid (IAA) 2 g/l 0.1 mg/l
Cefotaxime 250 g/l 250 mg/l
Timetin 150 g/l 150 mg/l
Phosphinothricin (PPT) 4 mg/ml 500 μl
Plant Preservative Mixture (PPM)
1 ml/ml 1 ml/l
Table 4: Rooting medium
Rooting Medium
Compound Final concentration
Duchefa M0231 MS + B5 4,4 g/l
Sucrose 30 g/l
Agar 8 g/l
Buffers
We used a CTAB buffer (Table 5) for DNA extractions and a TAE buffer (Table 6) for electrophoresis gels.
Table 5: CTAB buffer
CTAB
Compound Concentration
CTAB 2% w/v
NaCl 1.4 M
2-Mercaptoethanol 0.2% v/v
EDTA 20 mM
TRIS-HCl 100 mM
Table 6: TAE buffer
TAE (1x)
Compound Concentration Concentration
Tris base 12.1 gl
Acetic acid 2.854 ml/l
EDTA 0.5 M
PCR compounds
The compounds used for PCR amplifications are detailed in Table 7. The primers used in either PCRs or sequencing are detailed in Table 8.
Table 7: PCR Compounds.
PCR compounds
Compound Stock concentration Final concentration
H2o 16.7 μl
Lab 10x buffer 10 x 1 x
MgCl2 50 mM 2.5 mM
dNTP 10 mM 0.1 mM
PrimerF 10 μM 0.4 μM
PrimerR 10 μM 0.4 μM
Taq 5 U/μl 0.06 U/μl
DNA 20 to 50 ng/μl 1.6 to 4 ng/μl
Table 8: Primers used for PCRs and sequencing.
PCR and sequencing primers
Gene Use Primer name Sequence 5’-3’
Cas9 PCR RO94 GGACACTTCCTCATCGAGGGT
RO95 GTGGAGCCTTGGTGATCTCGG
ERF024 Expression assay, PCR, sequencing
ERF24_F1 TAACCCTCCCCTTCTTCACA ERF24_R CGAATTGGGAAAGTTGAGGT
OFP1 PCR, sequencing (forward only)
OFP_g_PreFw GTACCCAATTTTCGGGGAGT OFP_g_PreRv ATCCAACGATGCCTACTTCG
Melon transformation
The transformation protocol used spans several days within a week, starting on a Monday as a day one.
It consists of two parallel procedures, the preparation of the Agrobacterium tumefaciens and the preparation of the seed explants, which converge when the explants are infected. On day 1 the Agrobacterium was plated with the plasmid on LB with the antibiotics spectinomycin and rifampicin and it was grown at 28 °C for 48 h. I did not perform this step.
On day 3 we disinfected the seeds (Figure 8). First of all, we removed the coat of the seeds with a scalpel and we performed a cut on the opposite part of the embryo. Then we bagged the seeds in a netting cloth closed with string in groups of 30 to 50 seeds. We hydrated the seeds in distilled water for two hours at room temperature. For the disinfection, we immersed the seeds for 2 min in 96 % ethanol and then 20 min in a solution of bleach (30 to 50%, 50 mL bleach + 50 mL water with 4 drops of tween 20). Afterwards, the bags were washed in sterile water 3 times for 10 min each.
Figure 8: Transformation protocol, day 3. Removal of the coat (left), and the jars where the disinfection takes place (right).
On another hand, also on day 3, we collected Agrobacterium from a single colony from the plates and used to inoculate 25 ml of liquid medium of LB spectinomycin and rifampicin. We kept the culture at 28
°C and 200 rpm overnight.
On day 4, we removed the remaining endosperm from the seeds in a laminar flow cabin using an stereoscope and we let the seeds grow in small petri plates with Germination Medium (Table 1) during 24h at 28 °C and long day conditions (16 h light/8h darkness) (Figure 9).
Also on day 4, we inoculated volumes ranging from 0.5 ml to 2 ml of the Agrobacterium culture into a solution of 50 ml of MB with 25 µl of Acetosyringone 400 mM. We prepared more than one flask each time to have more than one backup in case one of the cultures would not grow.
Figure 9: Transformation protocol, day 4. Removal of the integument (left) and letting the seeds grow in small plates (right)
On day 5 we measured the optic density at 600 nm of the Agrobacterium culture. If it was over 0.5, we diluted the culture assuming a direct relationship between optic density and concentration, to reach an optic density of 0.4-0.5. We poured a small volume of the culture into petri dishes and there we cut the explants from the seeds (Figure 10). Next, we placed the explants into the Agrobacterium culture for 20- 30 minutes. After that, we put the explants with the culture medium in a syringe, and we pulled the plunger five times to produce vacuum and improve infiltration of Agrobacterium. After that, we dried the explants with paper and we placed them in plates with Co-culture Medium (Table 2) until day 8 at the same temperature and light conditions (Figure 11Figure 10).
Figure 10: Cuts applied to the seeds. Since there are two cotyledons, we obtained 4 explants per seed.
Figure 11: Transformation protocol, day 5. Explants cut from the seeds (left), Agrobacterium infiltration (center), 12 explants placed in co-culture medium (right).
On day 8 we washed the explants with antibiotics (Cefotaxime at 250 mg/L and timetin at 150 mg/l), we dried them on paper and and we transferred them to Selection Medium (Table 3). Then we selected the transformed explants and the regeneration begun to regenerate callus.
Invitro regeneration and individualization
Every 10 to 15 days, we cleaned the calli: we discarded any part that was vitrified, necrotic by the selection herbicide, not green or stalled in regeneration (Figure 12). Also, if any callus had a contamination with Agrobacterium or other bacteria, we cleaned it with antibiotics and we transferred it to a new plate.
Figure 12: Maintenance of the callus. On the left callus left on selection medium for 10 days after last cleaning, on the right, the callus after removal of unwanted parts.
When the calli were of a certain size, we increased the height of the petri dishes and when they started to produce shoots, we transferred them to tubes with selection medium (Table 3). When the top two nodes were big enough, we transferred them to a tube with rooting medium (Table 4). At this stage, we collected the leaves from the bottom nods to determine whether the transformation and edition had worked and to check if the plant was diploid. After the rooting medium, the plant would be ready to acclimatize to be grown in soil.
Genomic DNA extraction
For the DNA extractions, the plant material available was either calli or leaves, which we collected in an Eppendorf tube. To begin the protocol, two methods were available to grind the material: placing 3 mm balls of tungsten or steel in the Eppendorf and using the mixer mill to grind after adding 500 µl of CTAB (Table 5) or freezing the tubes in liquid N2 and grind manually with a pestle before adding 500 µl of CTAB (solution 1). Secondly, we incubated the tubes at 65 °C for 30 to 45 minutes. Next, we added 500 µl of chloroform:isoamilic acid 24:1 solution and mixed in the tubes with the vortex, followed by a centrifugation at 11000 rpm for 10 min which precipitated the mixture and allowed the supernatant to be recovered and transferred to new tubes. Then, we transferred the supernatant to new tubes, adding 350 µl of isopropanol. We mixed the tubes by inversion and centrifuged them at 10000 rpm for 5 min, so that the supernatant could be discarded. We kept the pellet and we washed it with ethanol and centrifuged one last time, at 10000 rpm for 3 min. Last step was to discard the supernatant, let the ethanol dry and add 50 µl of ultra-pure water. We kept the tubes in a fridge overnight before they were frozen at -20 °C or used.
PCR
The common components for the PCR experiments, used to prepare a “mastermix”, are described in Materials (Table 7). Depending on the purpose of the PCR, we used a different combination of 1 µl at 10 µM of each of the specific primers (Table 8) and 2 µl of previously extracted DNA. As for the reaction cycle, it starts with a denaturalization of the DNA at 94 °C for 1 min, followed by 35 cycles of denaturalization at 94 °C for 30 s, annealing at the primers melting temperature for 30 s and elongation at 72 °C for a variable time, with a minimum of 30 s, depending on amplicon size (approximately 1 minute per kilobase), and finished with a 5 min elongation at 72 °C. We validated the performance of the reaction using a NanoDrop ND-1000 spectrophotometer and an agarose gel.
For colony PCR, we used 2 μl of a colony diluted in 50 μl of purified water as a template instead of genomic DNA.
Agarose gel for DNA electrophoresis
Depending on the number of samples, we chose different sizes of agarose gel so that the size of the tray and the comb matched the number of samples. To begin, we weighted the agarose. Then we added TAE 1X (Table 6) to an agarose concentration of 1 % w/v. The final volume depended on the desired size of the gel. We heated the mixture in the microwave oven to boiling point and then we cooled it until the flask could be manipulated. After that, under the gas extraction hood, we added 28 µl of Ethidium Bromide 0.07% per each 100 ml of TAE (final concentration of 200 ng/uL). We added the mixture to the tray, which had its open sides sealed with tape, and we removed the bubbles. We then let the gel to solidify.
Simultaneously, we filled a buffer tank that fitted the gel with TAE 1X. We mixed the samples (amplified DNA derived from a PCR) with loading buffer.
To load the samples, we placed the gel in the tank and we removed the tape and comb or combs. We pippeted each sample and the ladder (λ digested DNA or 50 bp commercial ladder) into a well. Next, we connected the loading tank to 100-120 V and the gel was allowed to run a suitable amount of time, depending on the size of the amplicon and the base pairs difference between the bands we expected to observe.
To analyze the gel, we placed it in an UV chamber, which allowed to see the samples under UV light and take pictures.
PCR product purification (Sepharose)
Before sequencing an amplified DNA sample, we removed some of the salts by a Sepharose column. To prepare the column, first we made a hole in 0.5 ml Eppendorf tubes using a heated needle. Then, we added glass beads (≈ 20 µl) and 6 times the volume of the sample to purify of the stock Sepharose. Next, we placed the Eppendorf tubes inside a 15 ml tube, not at the bottom of it, held by the lid and tape at the upper part. Then we centrifuged the tubes at 2000 rpm for two minutes and a white column would be observed. A scheme of the columns is shown in Figure 11.
Figure 13: Final assembly of the tubes for the purification.
To purify the samples, we placed the 0.5 ml tubes with the column inside 1.5 ml Eppendorf tubes which in turn, we placed in 15 ml tubes. Then, we added 10 µl of amplified DNA and after another centrifugation the purified DNA (free of salts) was held in the 1.5 ml tubes.
RNA Extraction
To extract RNA we used the “Spectrum Plant Total RNA Kit®” . It provides all solutions and filter columns needed and it requires 100 mg of plant samples homogenized to a fine powder. For that, we grinded the samples, stored at -80 °C, in a mortar with liquid N2. Then, the procedure specified in the kit manual started by adding 500 μL of the “lysis” solution to a tube with the tissue powder and an incubation of 5 min at 56 °C followed by a first 3 min centrifugation. After we collected the supernatant, pipetted into a filtration column and we centrifuged it. Next, we collected the flowthrough and we mixed it through pipetting with 500 μl of the “binding” solution. We then transferred the mixture to a different column twice. After that, we added the “wash 1” solution and we centrifuged the tube. We performed the same procedure for the “wash 2” solution, twice. After a final centrifugation to dry the column, we placed it in a collection tube. Then, we added the “elution” solution and centrifuged the tube that contained the purified RNA. Finally, we checked the concentration by spectrophotometry at Nanodrop, and the integrity of the extracted RNA was checked in an agarose gel, as detailed above.
Retrotranscription (RT-PCR)
To retrotranscribe the previously extracted RNA, we used a “PrimeScript® Takara kit” . First step of the kit was to hybridize the primer with the RNA. For this we prepared a solution of 1 μl of oligo dT (50 µM), 1 μl of dNTP mix (10 mM each) and 8 μl of water. An oligo dT is a single-stranded sequence of deoxythymine used as a primer for reverse transcriptase reactions. Water was ultra-pure and autoclaved. We then incubated the solution at 65 °C for 5 min.
Second and last step was the retrotranscription reaction. To the previous solution we added 4 µl of 5x PrimeScript® buffer, 0,5 µL of RNAse inhibitor (40 U/µl) and 1 µL of RT PrimeScript® enzyme (200 U/μ), together with 4.5 µL ultra-pure autoclaved water. We incubated the solution at 42 °C for 45 min, then we inactivated the RT enzyme by heating for a period of 15 min at 70 °C. The cDNA obtained was finally diluted in ultra-pure water. For the subsequent PCR, we carried out the same steps detailed in the regular PCR, only swapping the extracted DNA with 2 µl of the retrotranscribed cDNA.
Results and discussion
Approach
The approach to validate the genes was to induce a knock out mutation on the gene to then observe variance in fruit phenotypes. For this we used CRISPR/Cas9 to edit the gene and cause a reading frameshift. A reading frameshift occurs when an indel (an edition, either insertion or deletion of bases) is not multiple of three and thus instead of deleting or adding an amino acid to the protein, it is read differently. When an edition causes a reading frameshift, the protein is completely different and the knock out is successful. When a plant with a reading frameshift in one allele is obtained, it can then be selfed to obtain plants homozygous for the edition. The fruit phenotype of this F1 plant can then be compared to the wild type to extract conclusions. However, it is possible to obtain a biallelic edition. In this case both copies of the gene would be edited, and if both caused a reading frameshift there would be no need for an F1.
Obtention of OFP1 edited plants and sequence analysis
The hypothesis is that OFP1 is the underlying gene of the shape effects caused by the QTL FSQS8.1 described in Díaz et al. 2014. To validate the function of OFP1, the ideal genotype to perform CRISPR/Cas9 edition on would be a Piel de Sapo introgression line carrying an introgression of PI124112 (Calcuta) containing the OFP1 gene: CALC8.1. This line yielded round fruits and the hypothesis is that by knocking out OFP1, it would recover the elongated fruit shape of PS. Díaz et al. 2014 described CALC8.1. However, transformation efficiency depends much on variety and previous attempts showed that the melon transformation protocol is significantly less efficient in PS than in Ved. Castelblanque et al., 2008 also reported that PS was much less efficient in transformation than Ved. Thus, we first attempted the experiment in Ved, which is a rounded fruit variety, with the hypothesis that OFP1 mutated individuals would have a longer fruit shape.
To validate the function of the OFP1 gene, we transformed embryo explants from a Ved pure line with Agrobacterium containing the CRISPR/Cas9 system to specifically edit this gene. We used two different plasmids, each with a different guide sequence, denominated guide 1 and guide 2. Both were within the gene, just 48 bp apart. The gene sequence with all relevant loci marked is shown in Figure 14. We tested two guides because in previous CRISPR/Cas9 experiments using different guide sequences gave different results.
Figure 14: Forward sequence of an exon fragment of OFP1. The first bases in pink are complementary to the reverse primer, and the latter complementary to the forward primer. These primers are detailed in table 8. The bases highlighted in blue are
complementary to the guide sequence used for CRISPR. Guide 2 is downstream of guide 1.
The transformation took place in the last week of November 2020, transforming a total of 400 explants (200 for each guide). For the guide 1, a bacterial contamination reduced the growth of the explants and all calli were discarded; on the other hand, for the guide 2, 81% of the 200 initial explants survived the transformation and formed calli. I joined the project 4 months after this transformation. The calli did not differentiate to shoots and leaves until May 2021. Some parts of the calli were necrotic or grew hyperhydrated instead of regenerating any shoot, so this required a continuous maintenance effort until regeneration started. During June 2021, we individualized 26 plants (Figure 15). All of them had been transformed with the guide 2 plasmid.
Figure 15: Individualized plants. On the plates, calli cultured on selection medium (Table 3). On the yellow cap tubes, individualized plants cultured on rooting medium with antibiotics (cefotaxime and timetin) (Table 3). On the blue cap tubes,
individualized plants cultured on rooting medium (Table 4).
We extracted DNA from these 26 individualized plants. We confirmed that 18 had the Cas9 gene via PCR, by detecting the amplification of a 454 bp specific sequence of the Cas9 gene (Figure 16).
Figure 16: PCR electrophoresis results for Cas9 in plant samples. Confirmation of the presence of Cas9 in 17 individualized plantlets. All samples had a band at ≈ 450 bp coinciding with the amplicon’s size.
Next, we sequenced 17 of these plants to check if OFP1 had been successfully edited and to identify the editions. We obtained the sequencing chromatograms and all had enough quality to be aligned to a reference Ved sequence except for plant 15. Mutated plants showed double peaks and discrepancies with the reference sequence as shown in Figure 17. In the wild types, no double peaks appear because the
genotypes used are pure lines, meaning they are homozygotic for all genes. When CRISPR causes an edition in only one of the alleles or a different edition in each allele, two different sequences appear overlapping which causes the double peaks. However, if the two alleles are edited identically, to discover the edition the chromatogram must be compared to the reference sequence. An example of this is found in the ERF024 experiment discussed below, also shown in Figure 23.
These mutations were found in two different loci. Plants 2, 5 and 10 had editions situated at 3 bp from the PAM sequence while the rest were edited at 70 bp from the PAM sequence. This can be clearly observed in Figure 18. In it, the sequences of plant number 10 and the wild type are aligned. The editions at 70 bp coincide with the guide 1 and it is possible that due to a mislabeling, those plants had the guide 1 construct instead. It could also be that the edition was slightly off target or that through the repairing mechanisms the edition happened 70 bp upstream.
Each sample had more than one sequence, because of the heterozygosity nature of the edition. I analyzed the sequencing results through Synthego ICE analysis, available at https://ice.synthego.com (Synthego Performance Analysis, ICE Analysis, 2019). This is an online tool used to analyze data from sequencing. It is useful because it shows data from each different sequence observed in a sample. Also summarizes data very well. Using this tool, I obtained the indel % and the KO-score. Indel % is the percentage of sequences of each sample that contained an indel. And the KO-score is the proportion of those indels that caused a reading frameshift. In our experiment, all plants had at least some edition present, ranging from 11 % to 97%. Plants with a low indel % could be wild type, those with an indel close to 50 % could be considered heterozygous while those with a higher percentage are probably edited in homozygosis. Regarding the KO-score, plants 2 and 5 had 75 % and 74 % respectively. They could potentially have biallelic frameshift editions and thus save time and effort of generating an F1. Plants 10, 11, 12 and 16 seem to have heterozygous frameshift editions. This presents a good prospect for CRISPR/Cas9 in melon as it shows a high edition efficiency and several plants are viable to continue the experiment, including two promising individuals that would not need an F1. However, 8 plants had an ambiguous indel %, hinting at the presence of somaclonal variations or chimeras. The indel % and KO-scores are shown in Table 9: Details of the sequencing results.Table 9.
Figure 17: Chromatograms from a section of OFP1 sequence. Plant number 10 and the reference sequence are aligned. The fragment includes the sequence complementary to guide 2. Each peak corresponds with its base in the sequence above. Green
highlighted bases in the sequence indicate double peaks or ambiguous readings. Red lines between bases represent deletions relative to the reference. Numbers under the sequence indicate the distance from the forward primer. Base 240 of the figure
corresponds with base 26974869 of the chromosome 8, according to the melon genome assembly v4.0 It is available at https://www.melonomics.net
Table 9: Details of the sequencing results. Indel % is the percentage of sequences in the sample that had a deletion. KO-score is the percentage of those deletions that would cause a frameshift mutation. Obtained by online resource available at https://ice.synthego.com (Synthego Performance Analysis, ICE Analysis, 2019)
In melon, during tissue culture, chromosome duplication is common. Tetraploid melon plants grow flatter fruits that may be viable, but in order to validate the gene, the mutation would need to be homozygous on all four chromosomes. This is not viable. Thus, we needed to know whether the individualized plants were diploid or not. A cytometry was performed in an external service and unfortunately, all 26 plants were tetraploid. Therefore, we deemed the experiment failed and we discarded the remaining plants.
More plants are still being individualized in search for diploid edited plants. A hypothesis is that since chromosome duplication is spontaneous in melon tissue culture, the extended time spent as a calli was causing the high percentage of tetraploidy. Genotype could also be a factor affecting tetraploidy. This is an important issue that needs to be solved to improve CRISPR/Cas9 efficiency in melon.
Figure 18: Aligned sequences of OFP1 from all plants. Highlighted in blue are the bases complimentary to guide 2. Highlighted in green are the bases with double peaks or ambiguous readings.
Plant Indel % KO-Score (%)
P1 11 9
P2 97 75
P3 23 0
P4 25 0
P5 97 74
P6 34 0
P7 40 19
P8 25 0
P9 33 2
P10 91 43
P11 73 48
P12 56 37
P13 21 0
P16 60 39
P17 23 0
P18 40 5
We also attempted the transformation twice in CALC8.1. However, both attempts failed due to a contamination that appeared during the period (72 h) of cocultivation with Agrobacterium (Figure 19). In the first transformation, once we observed the contamination was observed, we cleaned 106 of 299 explants and transferred them to selection medium. They did not survive. On the second transformation we did not use the syringe to favor the Agrobacterium infiltration but still all explants were contaminated and none survived.
We hypothesized that it could be that the Agrobacterium had overgrown and we tested it by colony PCR using the primers to amplify the Cas9 gene (Figure 20). Since a Cas9 band did indeed appear, at least Agrobacterium was present in the infections supporting the hypothesis of Agrobacterium. To confirm these results, we performed a regeneration assay with embryo explants obtained through the same method than in the transformation, but without the Agrobacterium infiltration. We cultured these embryo explants in selection media without phosphinothricin (PPT) nor antibiotics. The mediums also had either 0, 1 or 2 ml/l of Plant Preservative Mixture (PPM). No contamination appeared in any of the plates of the assay, supporting the overgrown Agrobacterium hypothesis. These explants, though they were no transformed nor infected still failed to regenerate into calli, showing the recalcitrance of the PS genetic background. In later essays, it was discovered that although Agrobacterium was present in the explants, the contamination was caused by an endogenous bacteria present in some of the seeds.
Figure 19: Contamination that appeared in the second attempt at transforming CALC8.1.
Figure 20: Electrophoresis results of the PCR for Cas9 in contamination colonies. Electrophoresis results of the PCR confirming the presence of Cas9 in the contamination colonies, and thus the presence of Agrobacterium. Samples of the “liquid” wells were
from the parts of the infection that were more liquid while the samples in the “solid” wells is from the solid-most parts of the infection.
To overcome this problem, an optimization of transformation and regeneration from PS genetic background is being carried at CRAG. Meantime, the validation of OFP1 gene depends on the identification and selection of diploid edited plants in Ved genetic background.
ERF024 transformation experiments
We took the same methodological approach for the ERF024 validation. We wanted to induce a knockout of the gene in a Vedrantais line using CRISPR/Cas9. The hypothesis is that ERF024 is the gene underlying the promotion of ripening in the QTL ETHQV8.1, reported Pereira et al. 2020. We also used two guide sequences for the same reasons explained in the previous section. A map of the relevant fragment of the gene is shown in Figure 21. One objective of this project was to individualize a plant with a knock out edition.
The first transformation took place on October 2020 and 80 % of explants survived and evolved into calli.
Some plants were individualized in January 2021 but they were not edited. I joined the project in March 2021, and these calli did not differentiate to shoots and leaves.
Figure 21: Forward sequence of an exon fragment of ERF024. The first bases in pink are complementary to the forward primer, and the latter complementary to the reverse primer. These primers are detailed in Table 8. The bases highlighted in blue are
complementary to the guide sequence used for CRISPR. Guide 2 is downstream of guide 1.
We confirmed that the calli had incorporated the plasmid via PCR test performed on DNA extracted from 4 calli, using specific primers for the plasmid (Figure 22). We also amplified the ERF024 gene and we confirmed its presence. Once we knew that the vector had been correctly transformed into the calli, the amplified ERF024 sequences were then purified and sequenced to identify if they were edited. The four callus were named g1-1, g1-2, g2-1 and g2-2. The first number refers to the guide they were transformed with. g1-1 had a deletion of one nucleotide in one allele and a deletion of two nucleotides in the other in the same position. g1-2 had the wild type allele in both chromosomes. g2-1 had an homozygotic deletion of 4 bp. g2-2 had an heterozygotic 4 bp deletion. A part of the chromatogram showing these editions is shown in Figure 23. Three out of four calli were edited, all editions cause a reading frameshift and two calli have biallelic editions. These are good results in terms of edition efficiency of CRISPR in melon. Also, preliminarily, guide 1 seemed to be slightly more efficient than guide 2. Only one allele out of four from the two g2 calli was wild type, versus two wild type alleles in the calli transformed with the guide 1. Though two samples per guide is not enough data to draw conclusions.
Figure 22: PCR electrophoresis results for Cas9 and ERF024 in calli samples. For both genes the g1-1 and g1-2 columns had amplified DNA from a callus transformed with the guide 1 plasmid, and the g2-1 and g2-2 columns had amplified DNA from a
callus transformed with the guide 2 plasmid. Rightmost column was the negative control.
