Crispr/Cas9 Mediated Gene Editing for the Improvement of Beef and Dairy Cattle

CRISPR/CAS9 MEDIATED GENE EDITING FOR THE IMPROVEMENT OF BEEF AND DAIRY CATTLE

KATHRYN POLKOFF

THESIS

Submitted in partial fulfillment of the requirements for the degree of Master of Science in Animal Sciences in the Graduate College of the University of Illinois at Urbana-Champaign, 2017

Urbana, Illinois

Master’s Committee:

Professor Matthew B. Wheeler, Chair Professor Jonathan E. Beever Professor Walter L. Hurley

Abstract

With the growing demands on agricultural production around the world, it is becoming increasingly important to figure out how to feed the growing population. For thousands of years we have been optimizing our agricultural outputs in livestock species through reproductive selection based on favorable traits, many of which are due to genetic variation. Single nucleotide polymorphisms are responsible for a great deal of either beneficial or undesirable mutations. In this study, we examine how we can use the CRISPR/Cas9 system to induce single-base pair substitutions, thereby improving animal genetics through removal of undesirable or insertion of beneficial mutations. In Angus beef cattle (Bos taurus), a mutation in the NHL-repeat containing 2 gene causes a heritable abnormality referred to as Developmental Duplications

(DD). Calves homozygous for this mutation are affected with a broad range of neural tube defects and commonly exhibit polymelia, which is the presence of additional limbs. The mutation has been identified as a single-nucleotide polymorphism resulting in a valine to alanine substitution in a highly conserved protein-coding region of the gene. Similarly, in Holstein dairy cattle (Bos taurus) a guanine to adenine substitution mutation in the α-lactalbumin flanking region at (+15) base pairs from the transcriptional start site has been shown to be associated with increased quantity of milk production, likely due to upregulation of the α-lactalbumin protein and increased lactose synthesis. Other breeds of cattle used for dairy production, such as Nelore and other tropically adapted Bos indicus breeds, do not have this mutation unless they are crossed to Holsteins. In this study, the CRISPR/Cas9 system was paired with a guide RNA targeting the aforementioned genes to create a break in the DNA of fibroblast cells from Nelore and Angus cattle. Cells were also supplied with a 90 nucleotide single stranded-oligonucleotide repair template, which allows for homology directed repair to introduce the desired substitution

at those loci. Cells from these cattle that are genetically improved by this technique may be used for somatic cell nuclear transfer and resulting embryos will be transferred to recipients. With these genetic improvements, we can preserve high quality beef cattle genetics while eliminating a harmful mutation, and we can increase milk yield of tropically adapted dairy cattle to feed malnourished populations around the world.

iii

Acknowledgments

First and foremost, I would like to thank my adviser Dr. Matt Wheeler for “recruiting me out of high school.” Thank you for bringing me into an environment that cultivated my passion for science, developed me as a researcher and challenged me to see the bigger picture, and for teaching me the importance of a good beer every once in a while. I’d also like to thank Marcello

Rubessa, who was with me from the first time I held a pipet, through two degrees and a whole bunch of Italian coffees. I am lucky to have been mentored by these two and numerous other professors, post-docs, graduate students, and researchers. Thank you Dr. Beever for always being around for a good conversation, good advice, and a good laugh—and for letting me pretend like I was in your lab when I needed to. Thank you Dr. Hurley, Dr. Knox, Dr. Miller and all of the amazing Animal Sciences faculty and staff; I relied on your expertise, your graduate students, and your smiling faces in the hall for years, and every day you make me thankful that I chose

Animal Sciences as my home. Thank you to all the brilliant graduate students and undergraduates I had the pleasure of working with, whether it was collaborating, brainstorming, or reassuring me that I’m not the only one who had gone entirely bonkers at some points. Extra thanks to Sammi Lotti, Rebecca Winters, and Lidia Sbaraini-Arend, who are absolutely irreplaceable, beautiful, brilliant, and hilarious. I also would like to thank the Jonathan Baldwin

Turner fellowship; none of this would have been possible without its support. Finally, thank you

Mom and Dad for teaching me how to be a human, then for pushing me to be the best human I can be. Thank you to my brother Michael, my friends, my roomies, and my teammates and coaches, all who played a special part in my time here. Once an Illini, always an Illini. I-L-L…!

Table of Contents

List of Abbreviations ...... vi

Chapter 1: Review of Literature ...... 1

Chapter 2: CRISPR/Cas9 mediated repair of NHLRC2 locus in Angus beef cattle ...... 19

Chapter 3: CRISPR/Cas9 mediated editing of the α-lactalbumin locus in Nelore dairy cattle .... 36

Chapter 4: General Discussion...... 55

References ...... 58

Appendix A: Additional Figures ...... 64

Appendix B: Protocols for Transfections and Analysis ...... 67

List of Abbreviations

Cas9: CRISPR associated protein 9 CRISPR: clustered regularly interspersed short palindromic repeats DD: developmental duplications GFP: green fluorescent protein gRNA: guide RNA HDR: homology directed repair LALBA: α-lactalbumin gene NHEJ: non-homologous end joining NHLRC2: NHL-repeat containing 2 NTD: neural tube defect PAM: protospacer adjacent motif PCR: polymerase chain reaction RFLP: restriction fragment length polymorphism SNP: single nucleotide polymorphism ssODN: single-stranded oligonucleotide TALENS: transcription activator-like endonucleases ZFN: zinc finger nucleases

Chapter 1: Review of Literature

The population of humans on earth has been increasing exponentially, and in 2050 it is expected to be 9 billion people. Hunger already plagues many regions of the world, and unless we take action, the number of hungry people will only continue to grow. Furthermore, as third world countries continue to develop and become more affluent, the demand for meat and animal products rises as a larger portion of the diet. Therefore, it is extremely important that we optimize the efficiency of our livestock production in order to make the most animal products using the fewest resources.

Because of the growing population and expansion of urban terrain, the amount of land used for farming will not be able to increase, yet we will need to increase meat production from

270 million tons annually to 470 million tons annually to meet demand by 2050 (FAO, 2016).

This means that we cannot simply upscale current production practices; we instead must continuously improve our practices and use fewer resources. Over recent years, we have already begun to optimize the conditions for nutrition, management, and reproduction practices.

Since the domestication of animals for livestock, humans have been selecting animals with desirable traits and thereby directing the evolution of the species. Some of these traits have included: temperament, fat content, size, growth rate, and feed conversion efficiency (Haskell,

Simm, & Turner, 2014; Koch, Swiger, Chambers, & Gregory, 1963). The best and most valuable animals have been bred with each other over generations, and high quality animals are selected to be the pool of genetics for the next generation of offspring.

Current reproduction methods such as artificial insemination, in vitro fertilization, and embryo transfer allow for a very small number of animals to be the majority genetic contributors

of the population. The widespread dissemination of superior genetics with artificial reproductive technologies has contributed heavily to the increase in production in recent years (Hansen,

2014). Additionally, in 2008 the FDA approved the sale of cloned animals and their progeny for food and production purposes, leading to yet another reproductive technique available to preserve and propagate advanced genetics (FDA, 2014).

1.1 Genetic mutations in cattle

While there are many advantages to these reproductive methods, there are a few drawbacks. One prominent disadvantage to heavy individual contributions to the gene pool is that recessive genetic disorders emerge more readily among the population. The elite animals are bred with each other with the aim of creating animals homozygous for superior traits, however they simultaneously can become homozygous for deleterious alleles (Van Eenennaam, 2014).

Most of these harmful alleles have emerged from mutations; errors in DNA replication have been the driving force for evolution, leading to immense diversity among living things over time. Even the smallest changes in the genetic code can have a large impact on an organism and its offspring. This can be beneficial as shown by evolutionary selection for favorable traits, but it can also have detrimental impacts. With respect to agricultural animals of high production quality, a mutation that goes unnoticed can become prevalent throughout the population within a matter of a few generations.

Some examples of mutations, specifically in cattle, are: bovine chondrodysplastic dwarfism (Yoneda et al., 1999), syndactyly (Drögemüller & Distl, 2006), arthrogryposis multiplex or “curly calf syndrome,” neuropathic hydrocephalus, and developmental duplications defect (DD) (Van Eenennaam, 2014). One example of how widespread these defects can become is demonstrated in dairy cattle; the bull Pawnee Farm Arlinda Chief, whose daughters showed 2

significantly increased milk production, is expected to have sired 14% of the current US dairy population. However, he was a carrier for a single nucleotide polymorphism (SNP) nonsense mutation in the APAF1 gene that causes truncation of a vital protein that causes fetal death if inherited in a homozygous fashion. Because of artificial insemination using his semen and that of his offspring, the trait has become so widespread that it is estimated to have been responsible for

525,000 abortions and costed the industry over about $400 million (Adams et al., 2016).

Conversely, some of these mutations are responsible for favorable production traits, such as the slick gene that decreases heat stress, a myostatin knockout for increased muscling, calpastatin for increased meat tenderness, a polled gene for lack of horn development (Casas &

Kehrli, 2016), and a single base pair substitution in the alpha-lactalbumin gene for increased milk yield (Bleck & Bremel, 1993a). There are 233 known Mendelian traits or disorders known in cattle, 133 of which have a known quantitative trait locus in the DNA (Nicholas, 2017).

1.2 Developmental Duplications

Developmental Duplications (DD) was first detected in Angus beef cattle on account of an increase frequency of malformed calves, especially those with polymelia—the presence of extra limbs (Denholm, Martin, & Denman, 2011). As a result, DNA from mutant calves was collected for a SNP chip analysis. A genome wide-association study showed that calves born with this defect consistently harbored the same SNP, which is a substitution mutation on exon 5 of the NHL repeat containing 2 (NHLRC2) gene. This SNP is a cytosine substitution for a thymine, which is a missense mutation in the protein coding region of the NHLRC2 gene. The wild type codon encodes a valine, while the mutant variant encodes an alanine, provoking an expected conformational change in a highly conserved region of the genome. The allele for DD is inherited in a Mendelian fashion, and heterozygotes carriers show no effects of the mutation.

Furthermore, the trait demonstrates incomplete penetrance, meaning that homozygous recessive individuals do not always show phenotypical abnormalities of DD (Denholm et al., 2016).

Because heterozygous individuals are unaffected, and in part owing to the widespread dissemination of genetics through artificial reproductive technologies, the allele frequency of this mutation grew to 7% in Australia and 3% in the USA before it was noticed (Denholm et al.,

2016). Furthermore, it is likely that the homozygous recessive genotypes are often embryonically lethal, and we therefore are unable to exactly quantify the costs of this mutation. In a genotypic study of Angus sires, the carrier frequency was 6% in the United States; however, there were no homozygous individuals found, supporting the theory that homozygous recessive genotypes frequently may be embryonically lethal (Beever, 2013).

1.3 NHLRC2 Gene

The Bos taurus NHLRC2 gene is located on exon 5 of chromosome 26 and is a protein coding gene with 12 exons and a total of 643 amino acids (NCBI Gene:3743554). The exact function of the gene is still unknown, but it has been shown to be expressed in alimentary, cardiovascular, nervous, olfactory, and visual systems in mice (Blake et al., 2017; Finger et al.,

2017).

Beyond the initially observed polymelia, calves with various neural tube defects have been confirmed to be DD recessive homozygotes. The phenotypic manifestations for DD include a range of neural tube defects such as: spina bifida and myemeningocoeles—sacs of spinal fluid protruding from the skull or spine due to lack of bone closure around the spinal cord, teratomas—solid tissue tumor masses, lipomas—fat filled tumor masses, diprosopus—facial duplication and abnormalities, polymelia—supernumerary limbs, and conjoined twins (Denholm et al., 2016). The majority of polymelia cases in these cattle are notomelia, meaning extra limbs 4

are found in the notochord region located along the dorsal midline, from the back of the head to the thorax (Denholm et al., 2011). These phenotypes indicate that this mutation in the NHLRC2 gene in bovine is a cause of neural tube defects (NTDs).

NTDs can be attributed to a variety of factors and cause many different conditions, but are unified in that they are a result of improper neural tube closure (Greene & Copp, 2014). In early embryonic development, after formation of the primitive streak, the neural plate extends from cranial to caudal ends of the embryo. The neural ectoderm then migrates inward, pushing together to form the neural fold and neural crest, followed by folding over the neural plate to form the neural tube. Apoptosis was shown to be present during neural fold bending and fusion, dorsal neural tube remodeling post neural tube closure, and migration of neural crest cells away from neural tube, though cell death is not necessary for neural tube closure itself (Massa et al.,

2009). Prevention of apoptosis in any one of these events could be the cause for neural tube defects. A live imaging study of mouse embryonic development by Yamaguchi et al. suggests that a balance of apoptotic and anti-apoptotic factors at specific intervals in development are necessary for successful neural tube closure. In this study, the absence of apoptosis induced via caspase inhibitors (zVAD-fmk) or apaf1 knockout in embryos agrees with the previous study in that the neural tube is able to close without apoptosis, however these embryos showed a significantly reduced neural tube closure speed. This delay in closure could increase the possibility for neural tube defects to develop during that sensitive time window (Yamaguchi et al., 2011).

To inhibit caspase activity, Yamaguchi et al. cultured embryos with the substrate mimetic caspase inhibitor zVAD-fmk, which is an irreversibly binding fluoro-methylketone synthesized to mimic the Val-Ala-Asp capsase binding substrate found in vivo, and showed reduced neural

tube closure time (Yamaguchi et al., 2011). In the amino acid sequence encoded by the NHLRC2 gene, a similar sequence, Tyr-Val-Ala-Asp (YVAD) is found three times in the coding regions of bovine NHLRC2, which is conserved in other species such as mouse and human (NCBI Gene ID:

534327). This YVAD sequence has been shown to be the binding site for caspase-1, and it has been shown that YVAD sequences are able suppress apoptosis by inhibiting caspase-1 or caspase-3-like proteases (Dorstyn, Kinoshita, & Kumar, 1998). While the DD associated mutation does not appear to affect the coding of the YVAD regions, alterations in the protein coding regions of NHLRC2 could change the shape of the protein and cause variation in anti- apoptotic properties of the protein.

Disruptions in cell death pathways have been shown to be detrimental during embryonic development. When the NHLRC2 locus was knocked out in a mouse model, it appeared to be embryonically lethal (Delhotal, 2016). Several studies show that knockouts or disruptions in the

Caspase 3, Caspase 8, Caspase 9, or Apaf1 genes, all of which encode proteins that are important for apoptosis pathways, are lethal in mice. Of those, mice with Casp3, Casp9 or Apaf1 knockouts show neural tissue malformation such as visible ectopic masses in their heads, disfigured cerebrum, exencephaly, and brain overgrowth (Hakem et al., 1998; Los, Wesselborg, & Schulze-

Osthoff, 1999; Meier, Finch, & Evan, 2000; Varfolomeev et al., 1998; Woo et al., 1998; Yoshida et al., 1998).

While neural tube defects and polymelia do occur in other animals, there have not been proven cases of DD in other species. A case of polymelia in a Holstein calf with similar phenotype to DD was found to be genetically unrelated (Neupane et al., 2017). However, in recent cases in a relatively genetically related ethnic group in Africa, there have been several occurrences of neural-tube related notomelia, in one particular case found in a child born to

genetically related parents. Due to the lack of DNA testing, no conclusive results can be determined from these cases with respect to mutation or genetic causes. However the NHLRC2 protein is highly conserved across species, and with the similar phenotypes shown in this population of humans as the Angus phenotypes, there is the possibility that there is a genetic locus for this trait on the NHLRC2 gene in humans (Kelani et al., 2017).

The NHLRC2 protein has recently been cited in human studies as a potential blood-based biomarker for Alzheimer’s disease (Long, Pan, Ifeachor, Belshaw, & Li, 2016), and furthermore as a potential hotspot for mutation in chronic myelomonocytic leukemia cells (Merlevede et al.,

2016). Further research must be done on this protein to uncover its function and its role in development.

1.4 Polymorphisms in the α-Lactalbumin Gene

In dairy cattle, studies are performed to establish quantitative trait loci and advantageous alleles for milk production. Several genetic variations at the α-Lactalblumin (LALBA) locus have been identified in Holstein (Bos taurus) dairy cattle. Three polymorphisms were identified by Bleck and Bremel in Holstein cattle in the 5’ region of the LALBA gene. Two of the mutations, at (+15) and (+21) base pairs 3’ from the transcriptional start site, were in the non- coding region of the mRNA. The third, at (+54), was identified as a silent mutation in the signal peptide coding region of the transcript (Bleck & Bremel, 1993b)

Further studies examined the correlation of the polymorphism at the (+15) position with increased milk yield in Holsteins; the single nucleotide substitution of a guanine to an adenine at this locus, (labeled as A, whereas the guanine variant is denoted with B), is positively correlated with milk yield. AA (P< 0.05) and AB (P<0.10) genotypes produced more milk than BB variants. The BB genotype was associated with higher fat and protein content, indicating that the 7

A allele likely causes the milk to be more dilute. The proposed mechanism for this is a higher lactose concentration, drawing in water to increase milk volume (Bleck & Bremel, 1993b). A study of allele frequency in 1993 indicated that at the (+15) locus, the A allele frequency was

32% in Holstein cattle, with 8.8% and 46.7% of the sample population of the AA and AB genotype, respectively (Bleck & Bremel, 1993b). Another study in 2008 examined allele frequency of the (+15) mutation amongst Holstein, Nelore, and Dairy Nelore populations.

Neither beef nor dairy Nelore sample populations harbored the A allele, while this Holstein sample population showed a 17.6% allele frequency. While this study showed that the (+15) mutation is fairly specific to Holstein cattle, it did not measure its impact on milk yield (Martins et al., 2008). A study in 2014 examined the impact of the (+15) LALBA mutation in Bos indicus

Sahiwal cattle in Pakistan. While there was no AA genotype found in any of the population, there were several individuals with an AB genotype, and an allele frequency for A of 7%. In this study, the AB genotype was highly correlated (P=0.07) with improved milk yield (Mir, Ullah, &

Sheikh, 2014). This agrees with the studies done by Bleck and Bremel in 1993 indicating that A has a positive impact on milk yield. With a larger population size or the presence of AA individuals, the correlation would likely have a higher significance. These studies show that AA and AB genotypes at the (+15) locus in both Bos Taurus and Bos indicus cattle improves milk yield. Furthermore, the allele frequency of A is much higher in Holstein, and that selection for or introduction of this polymorphism into Bos indicus could have a positive impact on milk yield.

1.5 Alpha-Lactalbumin and milk composition

The protein -lactalbumin encoded at that locus is a whey protein found in milk. It is one of six major milk proteins, and composes about 21% of the whey milk proteins, and 4% of total proteins in skim milk (Davies, Holt, & Christie, 1983). While about 80% of milk proteins are

caseins, α-lactalbumin and β-lactoglobulin are the major whey proteins. Beyond proteins, lactose is by far one of the most important components of milk for its roles both in milk synthesis and nutrition. Composing about 4.6% (w/w) of milk, and 36% of the dry matter in bovine milk, lactose makes up the highest proportion of the dry matter components (Walstra, 1999).

Research suggests that the sugar lactose is a major regulator of milk synthesis.

Membranes of the golgi, which is the site for lactose synthesis, are permeable to the monosaccharides glucose and UDP-galactose, which are the substrates for lactose synthesis.

However the membranes are not permeable to the disaccharide lactose, and therefore water is drawn in by osmotic pressure depending on the concentration of lactose. Studies have shown that a short term insulin infusion (decreasing blood glucose availability, therefore limiting a key substrate for lactose synthesis) was shown to decrease milk output with no effect on protein or fat yield (increasing the percentage), while the lactose concentration in the milk remained the same, supporting the theory that lactose is responsible for controlling milk volume (Holt, 1983;

Rook & Hopwood, 1970).

Lactose is composed of galactose and glucose subunits. In the cytoplasm, one glucose is converted to a UDP-galactose, and is actively transported into the golgi lumen along with unmodified glucose. The -lactalbumin protein is synthesized in the rough endoplasmic reticulum before being transported into the lumen of the golgi. In the lumen, β-1,4- galactosyltransferase, if complexed with α-lactalbumin, forms the lactose synthase complex, and catalyzes the production of lactose from glucose and UDP-galactose (Kuhn, 1983).

While β-1,4-galactosyltransferase is found in all tissues of the body, α-lactalbumin is tissue specific and is only produced in the mammary gland, and only during lactation. The α- lactalbumin content of the mammary gland increases slowly during gland development in 9

gestation, but increases greatly during lactation in order to yield the proper conditions for high milk synthesis (Bell, Beyer, & Hill, 1976; Brew, Vanaman, & Hill, 1968). While the lactose sugar is a key regulator of milk composition and volume, α-lactalbumin is the key regulator for lactose synthesis, and is key for controlling milk output.

1.6 Genome editing technology

Genome editing is the intentional modification of the DNA of an animal to elicit a desired phenotype (Esvelt & Wang, 2013). The first recorded instance of the creation of transgenic livestock was reported in 1985, when microinjection of a 2.6kb fragment of linear

DNA into the pronucleus or nucleus of pigs, rabbits, and sheep yielded expression of a transgene

(Hammer et al., 1985). It is of note that in this particular study, 5000 ova were injected, yielding

500 neonates or fetuses, 49 of which showed integration of the DNA, and 14 of which expressed the transgene. Since then, efficiency of integration has increased exponentially, and more importantly, we are able to target the location in the genome for DNA modification, as opposed to previous methods relying on random integration.

While microinjection of DNA can be used to create transgenic animals by random integration or low-efficiency homologous recombination, programmable nucleases allow for site- specific gene targeting at a much higher efficiency. Endonucleases can be programmed to target a specific area of the genome and induce a double-stranded break (Cox, Platt, & Zhang, 2015).

Repair of these double-stranded breaks is dependent on the endogenous mechanisms of either non-homologous end joining (NHEJ) or homology-directed repair (HDR). NHEJ occurs when the cleaved ends of the double-stranded break are ligated back together without the need for a repair template. NHEJ is referred to as “error prone” due to its tendency to cause insertions and deletions. Homology directed repair of a double stranded break uses an identical or quasi-

identical DNA molecule (such as the homologous chromosome) as a template for recombination, and is generally regarded as a faithful mechanism for recombination, provided that the template molecule is the same as the cleaved molecule (Pardo, Gómez-González, & Aguilera, 2009).

Hijacking of these endogenous repair mechanisms can lead to the knock-out (deletion), substitution, or insertion of DNA sequences or genes. NHEJ is used frequently to cause a frameshift or deletion that renders a gene unexpressed or “knocked-out,” while HDR can be used by supplying an exogenous repair template that contains regions of homology around the double- stranded break. A site-specific double-stranded break can be induced in the genome using programmable nucleases, such as zinc finger nucleases (ZFNs), transcription activator-like endonucleases (TALENs), and clustered regularly interspersed short palindromic repeats

(CRISPRs).

The first popular programmable endonucleases were Zinc Finger Nucleases (ZFNs).

ZFNs are a protein comprised of a specific zinc-finger protein (comprised of approximately ~30 amino acids), and a non-specific FokI endonuclease, which must dimerize with another FokI monomer in order to create an active nuclease. This requires two recognition sites which provides high specificity, however they are limited by the number of DNA target sites for this binding, and sometimes have low efficiency in cleavage (Durai et al., 2005). TALENs are similar to the ZFNs in that they also have a FokI endonuclease used to cleave the DNA, but they are instead tethered to transcription activator-like effector molecules, which are harvested from bacteria and can be designed to target any DNA sequence. TALENs received the Nature

Methods “Method of the Year” in 2011 for the impact they had on science (Joung & Sander,

2013; H. Kim & Kim, 2014).

The main drawback to ZFNs and TALENs is that they are difficult to design and synthesize, and can take a long time to be prepared. Recently, RNA guided endonucleases have emerged as a popular solution, specifically with use of the CRISPR/Cas9 system.

CRISPRs were first discovered by Francisco Mojico in 1993 when he was working on characterizing enzymes in microbes and he found an unusual group of palindromic DNA repeats.

He researched these structures and gene sequences for years and found similar sequences in other microbes, and eventually the acronym CRISPR evolved for the “clustered regularly interspersed short palindromic repeats” he had discovered (Lander, 2016; Mojica, Juez, & Rodriguez‐Valera,

1993). CRISPR and CRISPR associated protein-9 (Cas9), were discovered to be an adaptive immune system in bacteria against bacteriophages. The bacteria, such as Streptococcus pyogenes, incorporate sequences of viral phage DNA in to their genome, and then using an attached RNA-protein complex, the system recognizes and cleaves matching DNA sequences and render the phages incompetent (Cong et al., 2013; Garneau et al., 2010). Emerging for gene editing purposes in 2013, the CRISPR/Cas9 system quickly became immensely popular for gene editing because it is relatively easy and quick to design and synthesize, and works with a high efficiency and specificity (Kim & Kim, 2014). CRISPRs require a 20 nucleotide target sequence encoded in a guide-RNA (gRNA) molecule, which must be followed by a 3 nucleotide protospacer adjacent motif (PAM) sequence in the target DNA (Cong et al., 2013). CRISPR

RNAs, referred to as crRNAs, are the small spacers found in the DNA repeats that are transcribed into RNA to identify the complementary target site within the genome, while trans- activating CRISPR RNAs activate the Cas9 proteins to cleave the DNA; the two were able to be combined to one simple guide-RNA or gRNA (Jinek et al., 2012). The CRISPR gRNA complexes with the Cas9 protein to form an active DNA nuclease.

The CRISPR system can be introduced into cells or zygotes as plasmid DNA, mRNA, or protein. The plasmid DNA is transcribed and translated in the cell once introduced, while administered mRNA only must be translated into protein, and the protein can act directly on the

DNA once transfected or injected. Each method has varying half-life within the cell and varying cutting and repair efficiencies.

In recent and current studies, methods for gene insertion and editing are being improved and made more specific. Pairing a catalytically inactive (dead) Cas9 protein (dCas9) with a cytidine deaminase has recently emerged as a method of inducing single nucleotide edits without double-stranded cleavage. This system includes a dCas9, with catalytic activity restored in the

HNH domain to allow nickase activity, a cytidine deaminase, and a uracil DNA glycosylase inhibitor (UGI), and is paired with gRNA as above. The dCas9 paired with the gRNA guides the complex to the target site, deaminates any cytosine bases within the 5-8 base pair window of the target sequence yielding a uracil, then induces a nick on the complementary strand, while the

UGI inhibits the reversion of U to G. Upon repair of the nick, 15-75% of DNA is permanently repaired with the desired edit (Komor, Kim, Packer, Zuris, & Liu, 2016; Nishida et al., 2016).

Subsequent studies have improved the range of PAM sequences that can be used, have increased the specificity of the window, and have been shown to successfully edit plants and animals in vivo (Y. B. Kim et al., 2017; Shimatani et al., 2017; Zong et al., 2017). Base-editing systems offer distinct advantages; they have decreased off-target effects, they can safely be used without fear of deleterious or cancer-causing mutations, and they eliminate the need for a repair template.

The continued development of base-editing systems such as these will be important for the repair of single-nucleotide polymorphism genetic defects and deficiencies.

Beyond traditional homology directed repair, several strategies are being employed and developed for the insertion of genes. These include microhomology-mediated end-joining

(MMEJ) (McVey & Lee, 2008; Nakade et al., 2014), nuclease assisted vector integration (NAVI)

(Brown, Woods, & Perez-Pinera, 2016), and homology independent targeted insertion (HITI)

(Suzuki et al., 2016). Knock-ins using MMEJ involve using CRISPR or other endonucleases to create a double stranded break in DNA, and then supply a donor vector with short, microhomology (5-25 bp) on either end of an insert to induce an error-prone end-joining. HITI and NAVI involve using the CRISPR/Cas9 system to cut sites both at the targeted genomic DNA location and the donor plasmid, and use what is thought to be non-homologous end joining for the repair of the cut ends to insert the donor segment. These technologies have been shown to work in vitro and in vivo for the targeted insertion of genes, and can be developed further for uses in agricultural production and modern medicine to restore or augment gene functions without the need for a repair template or lengthy regions of homology.

1.7 Agricultural applications for gene editing

The aforementioned targeted gene editing systems are being employed worldwide for the purpose of enhancing livestock animal production. These genetic modifications can improve food quantity worldwide, improve animal welfare, and decrease the environmental footprint of livestock production.

Genetic engineering of livestock has occurred for years to enhance production and introduce desirable traits. For example, in 1989, AquAdvantage Salmon were first produced:

Atlantic salmon harboring a growth gene from a pacific salmon under the control of a promoter from a fish called an ocean pout. This allows these fish to grow all year long, enabling them to reach market weight in half the time required for the wildtype (FDA, 2017). In 1998, our lab

engineered pigs to produce an increased amount of milk for increased piglet growth and survival by insertion of the bovine alpha-lactalbumin gene (Bleck, White, Miller, & Wheeler, 1998). In

2005 dairy cattle were engineered to be resistant to mastitis from Staphylococcus aureus by expression of the human lysozyme enzyme (Wall et al., 2005). EnviropigsTM, a strain of

Yorkshire pigs, were engineered in 2001 using the murine parotid secretory protein promoter and signal peptide to express the enzyme phytase in their saliva. The phytase enzyme allows the pigs to digest previously biologically unavailable phosphorus from corn, barley and other cereal grains, which eliminates the need for supplemental phosphorus and decreases the phosphorus material excreted by 60% (Forsberg et al., 2003). In 2007, scientists knocked out the gene PRNP encoding cellular prion protein in Holstein cattle using homologous recombination, thereby preventing the transmittance of Bovine Spongiform Encephalopathy (Richt et al., 2007).

While engineered animals such as those above have been created for years, genetic engineering changed with the advent of targeted gene editing, which is defined as the use of site- directed nucleases to alter DNA at a predetermined location in the genome (Van Eenennaam,

2017). This allows us to introduce selectable traits without integrating recombinant DNA, to quickly integrate traits that we could select with traditional breeding selection over time, or simply to remove undesired genes. For example, in 2014, Ni et al used CRISPR/Cas9 to knock out several genes in goats; in their study, beta-lactoglobulin, a major milk allergen, was knocked out to decrease allergic properties of goat’s milk, prion protein was knocked out to prevent scrapie, a disease like BSE, and myostatin was knocked out for increased muscling (Ni et al.,

2014). In 2015, the myostatin gene was knocked out of sheep to increase the double muscling phenotype for increased meat production (Crispo et al., 2015). In 2015, cattle were engineered to be resistant to tuberculosis by a TALE-mediated SP110 gene knock-in at a targeted locus (Wu et

al., 2015). In a different approach in 2016, cows were engineered to be resistant to

Mycobacterium bovis infection by expressing the human β-defensin-3 (Su et al., 2016). In yet another approach in 2017, CRISPR-mediated targeted insertion of the NRAMP1 gene into the cattle genome was used to confer tuberculosis resistance(Gao et al., 2017). In 2015 pigs were engineered using CRISPR/Cas9 to be resistant to porcine reproductive and respiratory syndrome virus (PRRSV) by knocking out the receptor required for viral entry into cells, CD163

(Whitworth et al., 2016). In 2016, the POLLED allele was introduced into dairy cattle using

TALENs and homology directed repair to eliminate the need for dehorning (Carlson et al.,

2016). In 2016, using Zinc-finger nucleases, the RELA locus of domestic pigs was edited to be replaced with the same gene from warthogs, conferring resistance to African Swine Fever

(Lillico et al., 2016).

Biopharmaceutical production is another highly popular use of transgenic animals. Using the above technologies, animals such as goats, cows, and pigs have been engineered to produce recombinant and therapeutic proteins in their milk. Goats producing, ATryn®, human antithrombin-III, in their milk, and transgenic rabbits producing RuconestTM, a C1 esterase inhibitor, in their milk are examples of FDA-approved biopharmaceutical products from transgenic animals.

Gene editing in livestock can improve the quantity of output, quality of animal lives and environmental effects of farming practices. Spreading the genes for disease resistance could save thousands of animals and people from detrimental outbreaks. We no longer have to use years of breeding to select for the right combination of traits in an animal. It is even possible to edit DNA in such a manner that it is impossible to tell that it has been edited. This could allow us to grow

animals in previously unfavorable conditions and increase output of current animals, hence providing us with the opportunity to feed impoverished and starving regions of the world.

1.8 Methods for gene transfer

Using these guided endonucleases, we can make transgenic animals using several different methods. Some of the most popular reproductive methods for generating transgenic animals for gene-editing include somatic cell nuclear transfer (SCNT) of edited cells and pronuclear or cytoplasmic microinjection (Horii et al., 2014; Tang, González, & Dobrinski,

2015). SCNT, also referred to as cloning, was first performed to make Dolly in 1996 (Campbell,

McWhir, Ritchie, & Wilmut, 1996; Wilmut, Schnieke, McWhir, Kind, & Campbell, 1997). With this technique, it is possible to transfect somatic cells with the CRISPR/Cas9 system, verify exact

DNA sequences with desired changes, and use those edited cells for nuclear transfer to produce genetically modified embryos. Cytoplasmic microinjection involves the injection of any combination of CRISPR plasmid DNA, RNA, or protein encoding the information needed to target, knock out, correct, or insert the gene of interest directly into the cytoplasm of one-cell embryos, (Hai, Teng, Guo, Li, & Zhou, 2014; Horii et al., 2014). Each method has varying degrees of skill and difficulty associated, and different levels of success have been shown depending on the species for each method.

1.9 Limitations to Gene Editing

Approval from the Food and Drug Administration for the commercial consumption of

AquAdvantage Salmon was requested in 1995. This approval did not come until 2016, and soon after approval it faced more limitations. While the FDA did not deny the request in the end, they essentially denied it by over twenty years of inaction, and without going to market, it is difficult to maintain financial support for that length of time.

The regulations were initially in place because of the unknown: unknown location of

DNA integration, unknown effects of foreign genes, unknown effects of using a virus to insert genes that could cause cancer. Over the years we have been genetically modifying animals, we have shown that the risks are negligible. Furthermore, with RNA-guided endonucleases, we can target DNA, and make precise edits for the genetic improvement of livestock. As we uncover and delve deeper into the mechanisms of the genome we can take advantage of the information and harness it for production. The biggest limitation to this technology is public acceptance; the only way we can advance is if the public is willing to accept these animals as equally safe and nutritious as traditional animals. This is an opportunity to solve many problems of our times, if only we do not limit ourselves.

Chapter 2: CRISPR/Cas9 mediated repair of NHLRC2 locus in Angus beef cattle

2.1 Introduction

In beef production, efficient growth, high quality carcass composition, and consistent pregnancy and calving are all important traits that have consistently been improved upon by genetic selection over decades. To keep up with global demands for food and meat, this upward trajectory must be consistently improved upon to create the most quality meat with the fewest resources. While nutrition and management conditions do have a significant impact, preserving and pushing genetic quality of production animals is vital to advance production. This is further demonstrated by the monetary value placed on high quality animals and their offspring to be used as breeding stock.

Developmental Duplications defect (DD) is a congenital defect found in Angus beef cattle, and many bulls and cows of high genetic merit are carriers. Calves homozygous for this mutation are often born with a variety of neural tube related defects such as polymelia (the presence of one or more extra limbs), spina bifida, teratomas, diastematomyelia (splitting of spinal cord), craniofacial dysmorphogenesis, and more (Denholm et al., 2011). A genome wide association study determined that these defects are highly associated with a single nucleotide substitution in a coding region of the genome that elicits a valine to alanine change in the amino acid sequence.

For a long time, gene editing technology has been used for insertion or deletion of certain genes. Recent developments in site-specific endonucleases present the CRISPR/Cas9 system: clustered regularly interspersed short palindromic repeats (CRISPR) paired with CRISPR-

ASsociated-protein-9 (Cas9). Directed by a 20 nucleotide guide RNA (gRNA) sequence which binds to complementary genomic DNA that is immediately followed by a protospacer adjacent motif (PAM) of NGG at 3’ end, the Cas9 protein makes a double-stranded break in genomic

DNA. This double-stranded cleavage can be repaired via non-homologous end joining or homology directed repair, mechanisms which can be harnessed for site specific gene editing by adding single stranded DNA repair templates or double stranded vectors for insertion.

The aim of this study was to determine the potential of the CRISPR/Cas9 gene editing system as a tool for repairing the single nucleotide mutation associated with the DD phenotype, thereby restoring the DD free genotype. The long term goal is to use repaired cells taken from

DD carriers for somatic cell nuclear transfer; in this way we can preserve and propagate animals of high genetic merit while making progress toward eliminating this destructive mutation from the population.

2.2 Materials and Methods

Cell Culture

Bovine fetal fibroblasts were a gift from J. Beever (Dept. of Animal Sciences, UIUC).

Fetal fibroblasts were made from Angus beef cattle and were either homozygous or heterozygous for the DD mutant genotype at the NHLRC2 locus. Fetal fibroblasts were maintained in 1:1

Dulbecco’s Modified Eagle’s Medium (DMEM) (sigma D5648): Ham’s F10 nutrient mixture, supplemented with 10% fetal bovine serum, 0.01µg/mL bFGF, 100 U penicillin G sodium ml-1,

100 mg streptomycin sulfate ml-1, and 0.25 mg amphotericin B ml-1. Adult bovine dermal fibroblasts were isolated from ear notches taken from heterozygous carriers of DD. Skin and hair

were removed using a scalpel, and tissue was diced into 1mm cubes and plated for one week at

37C in DMEM supplemented with 10% fetal bovine serum, and 100 U penicillin G sodium ml-

1, 100 mg streptomycin sulfate ml-1, and 0.25 mg amphotericin B ml-1, until a fibroblast cell monolayer had been established. Both fetal and dermal fibroblasts were kept below 90% confluence and passaged using 0.25% Trypsin-EDTA (Fisher 25200056).

Plasmids and Repair Templates

Three 20 nucleotide gRNAs targeting the NHLRC2 locus within 50 base pairs of the DD mutation were designed according to the methods previously described (Ran et al., 2013).

Sequences with NGG or NAG PAM site were selected according to off-target and specificity score, evaluated using the E-CRISP design tool (Heigwer, Kerr, & Boutros, 2014). gRNAs oligos were annealed to form duplexes and ligated using 5’ CACC and 5’ AAAC overhangs as previously described (Ran et al., 2013) into gRNA/Cas9 dual expression vectors, either pSpCas9(BB)-2A-GFP (PX458) (Addgene plasmid # 48138), or pSpCas9(BB)-2A-Puro

(PX459) V2.0 (Addgene plasmid # 62988). Plasmids were a gift from Feng Zhang

(Massachusetts Institute of Technology). Plasmids were transformed into StellarTM Competent

Cells (Clontech); 10ng of plasmid DNA was incubated with 50 µl of competent cells for 30 min on ice, cells were heat shocked at 42C for exactly 60 seconds and placed back on ice for 1-2 minutes. 500 µl of LB were added and cells were incubated by shaking at 225rpm for 1 hour at

37C, and 100 µl were plated on LB agar plates with 100 ug/ml ampicillin. Well isolated colonies were selected, used to inoculate LB broth with 100 ug/ml ampicillin and shaken overnight at 225rpm at 37C. Plasmids were extracted using QIAGEN® plasmid mini kit. gRNA sequence with desired insertion was verified by sequencing using the human U6 forward promoter primer (LKO.1 5'GACTATCATATGCTTACCGT (Weinberg Lab)) to amplify the

insertion site. Custom single stranded oligonucleotide HDR templates were synthesized (IDT). ssODNs were 90 or 100 nucleotides with 45-50 base pairs on either side of the cut-site.

Nucleotides in the repair template were altered at the CRIPSR binding sites to prevent re-binding after HDR.

Transfections

Cells were seeded at a 75,000 cells per well 24 hours before transfection in a 6 well plate. One hour before transfection, media was replaced in well to DMEM and 10% FBS with no antibiotics. Cells were transfected with Fugene HD according to the manufacturer’s protocol.

1.5µg of DNA were combined with 5µl of Fugene HD in opti-MEMTM reduced serum medium and added to each well. 24-48 hours after transfection, GFP expressing cells were sorted using the Fluorescence-Activated Cell Sorting BD FACS Aria II at the Flow Cytometry Facility in the

Roy J. Carver Biotechnology Center at the University of Illinois. 5 cells per well were sorted into normal growth medium in 96-well plates, or 5,000 cells per well were sorted into a 6-well plate.

At confluency, cells were passaged to larger wells and DNA was extracted for further testing.

DNA extraction and evaluation

For genomic DNA (gDNA) extraction, cells were washed and pelleted and resuspended in 50 µl of QuickExtractTM (Epicentre), vortexed vigorously, incubated at 65C for 2 hours, denatured at

95 for 10 minutes, and stored at -20C until further analysis. PCR primers were designed to amplify a 435 bp segment surrounding the mutation on the NHLRC2 locus. PCR was performed using a 25µl reaction volume composed of 100ng genomic DNA, 12.5 µl Taq JumpStart

Readymix (Sigma P2893), 0.2 µM forward primer, 0.2 µM reverse primer, and ddH2O to 25 µl.

The reaction was placed in the thermal cycler and subjected to 5 minutes at 95°C; 30 seconds at

94°C, 30 seconds at 56°C and 30 seconds at 72°C for 30 cycles; 15 minutes at 72°C, and 5 min

at 10°C. MWOI enzyme is a site-specific restriction endonucleases that recognizes the sequence containing the SNP mutation indicative of DD (GCNNNNNNNGC), and cleaves the PCR product if the DD mutation is present in the DNA resulting in fragments of ~330bp and ~100bp.

1.5µl of 10x CutSmart Buffer (NEB), 0.5µl MWOI enzyme (NEB), and 8.0µl of ddH2O were added to 20µl of PCR product and incubated at 60C for 1 hour. Loading dye was added to each sample and 7µl were loaded into a 2% agarose gel containing 100ng/mL ethidium bromide.

Samples were subject to electrophoresis at 90 Volts for 30 minutes and visualized under UV illumination (GelReader). For purification for sequencing or subcloning, fragments were excised from a 1.0% agarose gel and purified using the Monarch® Gel Extraction Kit (NEB).

Subcloning and Sequencing

Cells were cultured and transfected as above. 24-48 hours post transfection, 2000-5000 fluorescent cells were sorted and pooled into one well of a 6-well plate. 3 days after the sort, cells were passaged with trypsin, washed, and gDNA was extracted. The NHLRC2 locus containing the defect was amplified using PCR as above. PCR products were cleaned-up

QIAquick PCR Purification Kit or gel-purified and amplicons were subcloned into pGEM®-T

Easy Vector System (Promega) according to the manufacturer’s instructions. Plasmids were transformed into DH5 competent cells; DH5α were stored at -80C, thawed on wet ice, incubated with 10 ng of plasmid for 30 minutes, heat shocked at 42C for 1 minute, placed back on ice for 1-2 minutes. 1 ml of LB was added and tubes were shaken for 1 hour at 37C; 200 µl were plated on agar plates containing 0.1mM Isopropyl thiogalactoside (IPTG), 60 µg/mL 5- bromo-4-chloro-3-indolyl-beta-D-galactopyranoside (X-Gal), and 100 µg/ml ampicillin. White and light blue colonies were selected using a sterile toothpick, shaken in 1 ml of LB broth with

100 µg/ml ampicillin at 37C for 20 hours overnight. Plasmids were extracted using PerfectPrep

BAC 96 Kit (Eppendorf). To verify presence of inserted PCR product, 2µl of plasmid DNA were added to 2.5µl of NEB CutSmart buffer, 0.5µl EcoRI Enzyme, and ddH2O to a total reaction volume of 25ul. Reaction was incubated at 37C for 1 hour. Products were subjected to electrophoresis on 1.5% agarose gel with 100ng/ml ethidium bromide and visualized with UV light. Colonies containing the insert showed two bands: one >2000 bp plasmid backbone and one

<400 bp DNA fragment excised from between EcoRI cut sites in the plasmid. 2µl of each well were added to 3.62 μL Sanger sequencing buffer (0.16 M Tris pH 9.0, 3.0 mM MgCl2, 4.9% tetramethylene sulfone and 0.001% tween 20), 0.25 μL BigDye® (Applied Biosystems), 0.08 μL

BigDye® dGTP (Applied Biosystems), and 1.00 μM reverse primer with a total reaction volume of 8.0 ul. The reaction was incubated in the thermal cycler for 90 seconds at 96C, then 54 cycles of 15 seconds at 96C, 15 seconds at 53C, and 3 minutes at 60C, followed by 10 minutes at

60C and a hold at 10C. Each reaction was purified via size exclusion using Sephadex G50 in a

25-30 MBPP Whatman Unifilter 96-well plate. After purification, the samples were sent to the

W.M. Keck Center for Comparative and Functional Genomics for analysis and Sanger sequencing was provided using a 3730xl DNA Analyzer. Resulting sequences were analyzed using CodonCode Aligner and BioEdit Sequence Alignment Editor.

2.3 Results and discussion

The aim of this study was to determine the potential of the CRISPR/Cas9 system for the repair of genetic defects in cattle. In this study, we chose to repair the single nucleotide mutation responsible for the developmental duplications defect in Angus beef cattle, an allele which has risen to a rather high frequency in recent years. A genome wide association study determined

that the DD phenotype was highly correlated with a TC non synonymous substitution in the

NHLRC2 gene on exon 5 of chromosome 26 in Bos taurus.

Our strategy was to use the CRISPR/Cas9 complex to introduce a double stranded break in the DNA in the region surrounding the NHLRC2 mutation associated with DD, and then supply a repair template harboring the normal genotype for homology directed repair. Cas9 complexes with a guide-RNA, which consists of a 20 nucleotide RNA sequence targeting the complementary genomic DNA sequence followed by the PAM, NGG. Three, 20 nucleotide gRNAs were designed, all targeting genomic DNA within 50 base pairs of the mutation. These three sequences, with cut sites 6, 28, and 42 base pairs away from the mutation site, targeted

DNA sequences of GATTGACCTAGAAGCTGAGA, TGGCATTGGAATTCAAGGTA and

CAAGGTACAGATAAAGAAGG, respectively (Figure 2.1). Each of these DNA oligos (IDT

Technologies) were cloned as a duplex into the pSpCas9(BB)-2A-GFP (PX458) (Addgene plasmid # 48138) plasmid encoding the gRNA scaffold and Cas9 from Streptococcus pyogenes

(Figure 2.2).

Transfection conditions for bovine fetal fibroblasts were optimized and found to be 6µl of

Fugene HD combined with 2µg of DNA with 75,000 cells in a 6 well plate. Bovine fetal fibroblasts homozygous or heterozygous for the mutation were transfected in these conditions with each of these plasmids and then sorted based on fluorescence and pooled in a 6 well plate.

Once cells recovered and reached confluence, they were analyzed using PCR and RFLP enzyme digest. Since the enzyme MWOI recognizes the exact sequence with the mutation present

(GCNNNNNNNGC), any disruption in the sequence prevents enzyme cleavage, whether it is the repaired genotype or an indel (Figure 2.3). GelAnalyzer analysis of the cleavage frequency compared to the controls indicated that guide-RNA (gRNA) 1 disrupted the mutation site at a

13.3% and 12.2% efficiency in cells heterozygous and homozygous for the mutation, respectively, while gRNA 3 disrupted the mutation site at 2.5% and 4.1% for homozygous and heterozygous cells, respectively (Figure 2.4). Minimal differences from the control were seen with gRNA 3 because it was 42 base pairs away from the cleavage site—not close enough to show any difference. Because gRNA 1 successfully targeted the optimal cut site at 6 base pairs away from the mutation, cleavage efficiency was not calculated for gRNA 2. For the remainder of the experiment, gRNA 1 was used.

To further verify the cleavage at the NHLRC2 locus, the PCR products were amplified and subcloned into pGEM®-T Easy Vector System (Promega), and subsequently sequenced. The sequence results support the enzyme cleavage data, showing that 16 sequences out of 76 had indels at the Cas9 cut-site, approximately 21.1%. The majority of the sequences showed deletions, with a few insertions (Figure 2.5). Once we knew the CRISPR was effectively cutting at the desired locus, we then designed a single stranded oligonucleotide (ssODN) repair template to supply for homology directed repair. This ssODN harbored the DD free “T” genotype to replace the C allele found in the mutant cells. Furthermore it contained a silent substitution mutation in the gRNA sequence to prevent CRISPR binding to the repaired genotype (Figure

2.6). Sequencing results from cells transfected with this system showed that 5 out of 60 sequences were correctly repaired, and that the repair efficiency is approximately 8.3% of edited cells (Figure 2.8).

From these data, we can conclude that the CRISPR/Cas9 system can successfully be used for the repair of congenital defects in cattle. To our knowledge, this represents the first time that this technology has been shown to be used for repair of congenital defects in cattle. To prove that this gene editing truly fixes the defect, future studies would need to be performed; for example,

generating a clone using the repaired genetic material from a homozygous recessive calf with phenotypic manifestations of DD. Nevertheless, the neural tube defects and possible embryonic loss associated with the homozygous recessive inheritance of the DD allele greatly decrease the value of these animals, and represent a great cost to the beef industry. Carriers of the defect are unaffected, so it was not until the carrier frequency was 6% in the US and 14% in Australia that the mutation was researched fully. The economic impacts of the relatively high frequency of this mutation are yet to be fully understood, especially because it is likely that the recessive homozygotes are embryonically lethal. However, to remove high quality bulls and cows that are carriers for this mutation from the population entirely would a) incur heavy costs against farmers who invested in these animals, and b) would remove beneficial alleles and quality genetics from the population along with this destructive allele. Providing a method to “fix” these cows and their progeny can resolve these problems.

Cloning is already a common practice among public and private farms. The FDA has declared meat and milk from cloned cows, goats and swine is as safe to eat as that of sexually produced offspring. Repairing the genetic defect would use the same procedures as cloning with the addition of a quick gene editing step to ensure the animal is no longer a carrier for the defect.

Cells would be taken from the animal of interest, cultured, repaired and fully sequenced, and then cloned and grown just as any other cloned animals. This system could be used not only for

DD in cattle, but also for a wide panel of genetic defects in livestock species. Gene editing paired with cloning allows farmers to promote the superior genetics of their high performing livestock while simultaneously restricting the propagation of destructive alleles. In this way, we can continue to advance the genetic quality of livestock to improve production quantity and maintain quality.

2.4 Figures

NHLRC2 in Bos taurus

Figure 2.1: Guide RNA target sequences on the NHLRC2 gene in Angus cattle. gRNA 1 (red), gRNA 2 (blue), and gRNA3 (green) all are followed by a PAM sequence of NGG or NAG. The DD associated SNP is denoted “mutant,” while the normal genotype is labeled “wildtype.”

Figure 2.2: Map of the spCas9 px458 plasmid into which each of the gRNA oligo duplexes were ligated.

Figure 2.3: RFLP enzyme digests of the NHLRC2 PCR products using enzyme MWOI. Lane 1) ladder, 2) empty, 3 and 4) Angus DNA heterozygous for the DD-associated mutation, 5 and 6) Angus DNA homozygous for the mutation, 7 and 8) Angus DNA homozygous normal without the mutation.

Figure 2.4: RFLP digests of PCR products using enzyme MWOI. Lanes 2-7 are amplified from DNA from Angus fetal fibroblasts with the indicated genotype, (Ho indicates homozygous for the DD associated SNP, Hz indicates heterozygous). Lanes 4-7 are pooled cell extractions, all transfected with the CRISPR/Cas9 plasmid from figure 2.2. Note that gRNA 1 shows noticeable difference in cleavage for both homozygous and heterozygous cells, indicating that the enzyme recognition site is no longer intact in some cells.

Figure 2.5: DNA sequences from heterozygous and homozygous Angus fetal fibroblast cells, transfected with the gRNA1 CRISPR/Cas9 plasmid, sorted and pooled. The red arrow indicates the location of the DD associated SNP. Various indels can be observed at the cleavage site (seven bases upstream of the mutation).

Figure 2.6: Sequence of the 90 nucleotide ssODN repair template transfected with the CRISPR/Cas9 plasmid. The red box labeled “1” indicates the position for the DD mutation, restored to the wildtype “T.” The red box labeled “2” indicates the silent mutation inserted to prevent CRISPR re-cutting correctly repaired DNA.

Figure 2.7: RFLP analysis with enzyme MWOI. Lane 1) ladder, 2) homozygous DD mutation control, 3-8) Pooled, sorted cells transfected with gRNA1 CRISPR/Cas9 plasmid and the ssODN repair template, originally homozygous for the DD associated SNP.

Figure 2.8: DNA sequences from edited cells transfected with gRNA1 CRISPR/Cas9 plasmid and the ssODN repair template. Uncleaved fragment from digestion in figure 2.7 was gel- excised, subcloned, and sequenced. Black arrow indicates site for DD associated SNP; “T” at that locus indicates that it has been correctly repaired to the normal genotype.

Chapter 3: CRISPR/Cas9 mediated editing of the α-lactalbumin locus in

Nelore dairy cattle

3.1 Introduction

With the global population expected to reach 9 billion people by 2050, we must produce

100% more food in the next thirty years to prevent a hunger epidemic worldwide (Simmons,

2011). We have worked for years to optimize production of crops, meat and dairy. The goal is simple: make more products with fewer resources. With great thanks to modern reproduction methods for improved genetics, milk production per animal has more than doubled in the past forty years. In dairy cattle, several studies (Bleck & Bremel, 1993a, 1993b; Martins et al., 2008;

Mir et al., 2014) have shown a correlation between a single-nucleotide polymorphism in the alpha-lactalbumin 5’ untranslated region at (+15) base pairs from the transcriptional start site and increased milk yield. This mutation is a guanine to adenine substitution in the non-coding region of the mRNA, and has been shown to be associated with increased milk yield in both Holstein and Nelore cattle, likely due to increasing the binding affinity of transcription factors. Since it originated in Holstein, the (+15) adenine allele is found at a high frequency among the Holstein population, but a very low frequency in the Nelore population and is achieved through cross- breeding.

The clustered regularly interspersed short palindromic repeats (CRISPR)/Cas9 system has been shown in recent years to be an effective tool for targeted gene editing (Cong et al.,

2013; Lillico et al., 2016; Ran et al., 2013). The CRISPR/Cas9 protein is paired with a guide-

RNA, which is an RNA loop that can be programmed to target any specific 20 nucleotide sequence within the genome, provided that it is followed by an NGG or NAG PAM sequence

(Jinek et al., 2012). The gRNA complexes with the Cas9 protein to induce a double stranded break in the three nucleotides upstream of the protospacer adjacent motif (PAM) sequence.

Mammalian cellular repair systems employ one of two mechanisms to mend these breaks: non- homologous end joining (NHEJ) or homology directed repair (HDR). By supplying an exogenous, single strand oligonucleotide (ssODN) repair template that has homology to the cut site, desired changes can be inserted at the target locus.

The goal of this study was to examine the potential for the CRISPR/Cas9 system to be used to edit the alpha-lactalbumin gene in Nelore cattle cells. The long term goal of this study is to use this technology for the targeted substitution of the adenine allele for the guanine allele in the alpha-lactalbumin promoter region, and use gene-edited fibroblasts for somatic cell nuclear transfer to create tropically adapted cattle with increased milk yield.

3.2 Materials and Methods

Cell Culture

Dermal fibroblast cells were a gift from C. Long (College of Veterinary Medicine & Biomedical

Sciences, Texas A&M University). Fibroblasts were isolated from skin of Nelore heifers. Cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) (sigma D5648) supplemented with 10% fetal bovine serum, 100 U penicillin G sodium ml-1, 100 mg streptomycin sulfate ml-1, and 0.25 mg amphotericin B ml-1 Dermal fibroblasts were kept below

90% confluence and passaged using 0.25% Trypsin-EDTA (Fisher 25200056).

Plasmids and Repair Templates

A gRNA targeting the LALBA locus with a cut site 5 base pairs from the desired mutation site were designed according to the methods previously described (Ran et al., 2013). The 20 nucleotide sequence GTGACCCCATTTCAGAATCT followed by the PAM sequence of TGG was designed as the gRNA target. DNAs oligos (IDT Technologies)

5’CACCGTGACCCCATTTCAGAATCT 3’and 5’AAACAGATTCTGAAATGGGGTCAC3’ were annealed to form duplexes and ligated into CRISPR plasmids using 5’ CACC and 5’

AAAC overhangs as previously described (Ran et al., 2013) into gRNA/Cas9 dual expression vectors, either pSpCas9(BB)-2A-GFP (PX458) (Addgene plasmid # 48138), or pSpCas9(BB)-

2A-Puro (PX459) V2.0 (Addgene plasmid # 62988). Plasmids were a gift from Feng Zhang

(Massachusetts Institute of Technology). Plasmids were transformed into StellarTM Competent

Transfections

For selection based on fluorescence, cells were seeded at 75,000 cells per well 24 hours before transfection in a 6 well plate. One hour before transfection, media was replaced in wells with

DMEM and 10% FBS with no antibiotics. Cells were transfected with Fugene HD according to the manufacturer’s protocol. 1.5µg of DNA were combined with 4.5µl of Fugene HD in opti-

MEMTM reduced serum medium and added to each well. Transfection reagent media was removed 24 hours later, and 24-48 hours after transfection, GFP expressing cells were sorted using the Fluorescence-Activated Cell Sorting BD FACS Aria II at the Flow Cytometry Facility in the Roy J. Carver Biotechnology Center at the University of Illinois. Five cells per well were sorted into normal growth medium in 96-well plates or 5,000 cells per well were sorted into 6- well plates. At confluency, cells were passaged to larger wells and DNA was extracted for further testing.

For selection based on puromycin, cells were seeded at 10,000 cell/well in a 24 well plate. A puromycin kill curve was performed for Nelore dermal fibroblasts; 1µg puromycin/ml media was determined to be the proper concentration for efficient selection. For transfection, 500ng of plasmid and 500ng of repair template in addition to 100ng of a puromycin resistance plasmid were added with 1.5µl Fugene HD per well. Cells were transfected as above. 24 hrs following transfection, media was replaced with normal growth medium. 48 hrs post transfection, cells were passaged, and 2 wells from a 24 well plate were combined to one 6 well plate containing 1 ug/ml puromycin. Media containing antibiotics was changed every other day for 5 days. After selection, cells were allowed to grow to confluence before DNA was extracted.

DNA extraction and evaluation

For genomic DNA (gDNA) extraction, cells were washed and pelleted and resuspended in 50 µl of QuickExtractTM (Epicentre), vortexed vigorously, incubated at 65C for 2 hours, denatured at

95 for 10 minutes, and stored at -20C until further analysis. PCR primers were designed to amplify a 443 bp segment surrounding the mutation at the LALBA locus, amplifying the region

(-363) to (+80) relative to the transcriptional start site. PCR was performed using a 25µl reaction volume composed of 100ng genomic DNA, 12.5 µl Taq JumpStart Readymix (Sigma P2893),

0.2 µM forward primer, 0.2 µM reverse primer, and ddH2O to 25ul. The reaction was placed in the thermal cycler and subjected to 5 minutes at 95°C; 30 seconds at 94°C, 30 seconds at 56°C and 30 seconds at 72°C for 30 cycles; 15 minutes at 72°C, and 5 min at 10°C.

PCR products were then subjected to a T7 endonuclease assay (NEB) to determine cleavage.

After amplification in the thermal cycler, products were annealed in the thermal cycler with an initial denaturation at 95°C for 5 minutes, ramped to 85°C at -2°C/second, then ramped to 25°C at -0.1°C/second, followed by hold at 4°C. 20µl of unpurified PCR annealed products were added to 0.5µl T7 endonuclease and 5µl NEB buffer 2 for a 50µl reaction volume and incubated at 37°C for 30 minutes. 2 µl of 0.5M EDTA was added to stop the reaction. 10µl were subject to gel electrophoresis on a 2% agarose gel containing 100ng/mL ethidium bromide and visualized under UV light.

PCR products were also subjected to restriction fragment length polymorphism (RFLP) analysis with a site specific restriction endonuclease. BpuEI enzyme is a site-specific endonucleases that recognizes the sequence containing the (+15) α-lactalbumin mutation (CTTGAG), and cleaves the PCR product if the mutation is present in the DNA. 1.5µl of 10x CutSmart Buffer (NEB),

0.5µl BpuEI enzyme (NEB), and 8.0µl of ddH2O were added to 20µl of PCR product and

incubated at 60C for 1 hour. Loading dye was added to each sample and 7µl were loaded into a

2% agarose gel containing 100ng/mL ethidium bromide. Samples were subject to electrophoresis at 90 Volts for 30 minutes and visualized under UV illumination (GelReader).

Colony Isolation

After transfection with GFP indicator and sorting by flow cytometry cells were seeded at low density (1000 cells/100mm diameter dish) in normal growth medium. After transfection with puromycin resistance plasmid, one well from a 24 well plate was passaged and seeded into a

100mm diameter plate in normal growth medium supplemented with 1ug/ml of puromycin, and subjected to antibiotic selection for 5 days. For both methods, after approximately 7-10 days tight colonies were visible in a microscope and marked on the underside of the plate. After media removal and rinsing with Dulbecco’s phosphate buffered saline (PBS), cloning cylinders were placed on the marked colonies using jeweler’s forceps. 1ml of 1% low melting point agarose heated and dissolved in PBS was pipetted around the cylinders and allowed to cool slightly before 40µl of trypsin was added on top of each colony within the cylinder as described

(Mathupala & Sloan, 2009). After 3-5 minutes, the trypsinized colonies were transferred to a 24 well plate where they were cultured in normal growth medium and allowed to reach confluence before analysis.

Subcloning and Sequencing

Cells were cultured and transfected as above. 24-48 hours post transfection, 2000-5000 fluorescent cells were sorted and pooled into one well of a 6-well plate. 3 days after the sort, cells were passaged with trypsin, washed, and gDNA was extracted. The LALBA locus containing the defect was amplified using PCR as above. PCR products were cleaned-up

QIAquick PCR Purification Kit or gel-purified and amplicons were subcloned into pGEM®-T

37C; 200 µl was plated on agar plates containing 0.1mM Isopropyl thiogalactoside (IPTG),

60ug/mL 5-bromo-4-chloro-3-indolyl-beta-D-galactopyranoside (X-Gal), and 100 µg/ml ampicillin. White and light blue colonies were selected using a sterile toothpick, shaken in 1ml of LB broth with 100 ug/ml ampicillin at 37C for 20 hours overnight. Plasmids were extracted using PerfectPrep BAC 96 Kit (Eppendorf). To verify presence of inserted PCR product, 2µl of plasmid DNA were added to 2.5µl of NEB CutSmart buffer, 0.5µl EcoRI Enzyme, and ddH2O to a total reaction volume of 25ul. Reaction was incubated at 37C for 1 hour. Products were subjected to electrophoresis on 1.5% agarose gel with 100ng/ml ethidium bromide and visualized with UV light. Colonies containing the insert showed two bands: one >2000 bp plasmid backbone and one <400 bp DNA fragment excised from between EcoRI cut sites in the plasmid.

2µl of each well were added to 3.62 μL Sanger sequencing buffer (0.16 M Tris pH 9.0, 3.0 mM

MgCl2, 4.9% tetramethylene sulfone and 0.001% tween 20), 0.25 μL BigDye® (Applied

Biosystems), 0.08 μL BigDye® dGTP (Applied Biosystems), and 1.00 μM reverse primer with a total reaction volume of 8.0 ul. The reaction was incubated in the thermal cycler for 90 seconds at 96C, then 54 cycles of 15 seconds at 96C, 15 seconds at 53C, and 3 minutes at 60C, followed by 10 minutes at 60C and a hold at 10C. Each reaction was purified via size exclusion using Sephadex G50 in a 25-30 MBPP Whatman Unifilter 96-well plate. After purification, the samples were sent to the W.M. Keck Center for Comparative and Functional

Genomics for analysis and Sanger sequencing was provided using a 3730xl DNA Analyzer.

Resulting sequences were analyzed using CodonCode Aligner and BioEdit Sequence Alignment

Editor.

3.3 Results and discussion

The aim of this study was to determine the potential of the CRISPR/Cas9 system for the introduction of a single base substitution in cattle. A GA substitution at the (+15) position 3’ from the α-lactalbumin (LALBA) transcriptional start site was selected as a target for editing because it is a mutation found naturally in Holstein cattle and is associated with increased milk yield (Bleck & Bremel, 1993a). At that locus, (+15) mutation (A allele) is theorized to increase the affinity of transcription factors for the α-lactalbumin milk protein, which is upregulated and in turn increases the production of lactose, the major osmole regulator in milk, and therefore increases milk volume (Brew et al., 1968). This mutation, is shown to be significantly correlated with increased milk yield (P<0.05 for homozygotes and P=0.07 in heterozygotes) in both

Holstein and Nelore cattle (Bleck & Bremel, 1993a; Mir et al., 2014). The CRISPR/Cas9 targeted endonuclease system provides an opportunity to introduce this mutation into tropically adapted dairy cattle for increased milk yield.

Here, we present a strategy for seamless substitution of the A allele at the LALBA (+15) locus in Nelore DNA. The CRISPR/Cas9 system is composed of Cas9 protein, which induces a double stranded break in the DNA, and a gRNA scaffold, which complexes to guide the protein to the cut site. In this study we used the pSpCas9(BB)-2A-GFP (PX458) (Addgene plasmid #

48138) plasmid to encode Cas9 and our gRNA scaffold, and cloned the 20 nucleotide target sequence 5’GTGACCCCATTTCAGAATCT 3’ into the scaffold backbone (Figure 3.1). This gRNA sequence binds to the (-8)to(+12) bases relative to the LALBA transcriptional start site, and induces a cut at between (+9) and (+10) (Figure 3.2). To insert the mutation using homology

directed repair, a ssODN of 90 nucleotides with 45bp homology arms on either side of the target site harboring the +15 single base substitution was supplied with the CRISPR/Cas9 system. In other studies, silent mutations are frequently added to prevent the rebinding and recutting of repaired DNA. In this system, since the target site is not a coding region and the effects of additional mutations would be unknown, the gRNA was designed so that the 3nt PAM sequence following the gRNA target sequence (NGG) would be altered by the GA +15 mutation (to

NGA), and thereby seamlessly prevents rebinding of repaired loci (Figure 3.3).

Dermal fibroblasts from Nelore cows lacking the mutation were used in the transfection.

Preliminary experiments optimizing transfection conditions showed that using Fugene HD at a

4.5ul:1µg DNA ratio in transfection conditions is optimal with minimal cell death, at a cell concentration of 10,000 cells/24 well plate (5250cells/cm2). The px458 plasmid encodes the GFP protein as a transfection indicator. While transfection with only the GFP plasmid or only the

CRISPR-GFP plasmid yielded transfection greater than 30% in adult dermal fibroblasts, the transfection efficiency as shown by GFP expression in flow cytometry was decreased to less than

1% of living cells when repair template was added. The amount of DNA was kept constant, and the amount of CRISPR plasmid was decreased as repair template ssODN was added. Due to low survival after flow cytometry, a puromycin resistance plasmid was added as a less stressful option, and a kill curve was calculated. The optimal concentration of puromycin for selection of

Nelore dermal fibroblasts was found to be 1µg puromycin/mL of media. After puromycin selection, wells were grown to confluency and DNA was extracted from the pooled collection of transfected cells.

DNA from pooled, transfected Nelore was amplified using primers for the LALBA flanking sequence, and the T7 endonuclease assay showed the CRISPR was successfully cutting

at the +15 locus (Figure 3). Pooled DNA was sent to sequencing and analyzed for decomposition to further verify cleavage at this locus using TIDE, and showed a 26.8% cleavage efficiency at the target locus (Brinkman, Chen, Amendola, & van Steensel, 2014) (Figure 3.5). The pooled

PCR products were then subjected to a restriction fragment length polymorphism digest using the BpuEI enzyme to determine whether the “A” allele was inserted at the (+15) position in the

LALBA gene. Because the BpuEI enzyme does not completely digest even the positive control, it is not possible to accurately predict the cleavage efficiency using PCR and RFLP alone.

Sequencing of the pooled PCR product confirmed that there was cleavage at that site (Figure

3.6). While there is a slight band visible from the pooled, transfected Nelore, the cleaved band does not align with the positive control band. These results show that this system successfully creates a double-stranded break 6 base pairs upstream of the (+15) site of the LALBA gene.

Further studies must be done to verify proper insertion of the A allele.

Dairy producers in temperate climates such as the United States have done a very good job optimizing milk yield, with Holsteins now producing on average 23,000 pounds of milk per animal per lactation (HolsteinAssociationUSA, 2017). The majority of this increase yield has been dependent on genetic selection and the advent of artificial reproductive technologies, such as artificial insemination and embryo transfer, in addition to improved nutrition and health status of the animals. Cows now produce on average double the amount of milk that they did forty years ago; it now requires 90% less land and 65% less water to make one gallon of milk than it did in 1944 (Simmons, 2011). This amount of milk is sufficient to distribute of milk, butter, cheese and other dairy products throughout the US, in addition to the export market growing yearly, with $5.3 billion of dairy products exported in 2015 (Newton, 2016). Nevertheless, it is estimated that 1 in 7 people around the world do not have access to enough nutrients, or suffer

from malnourishment (Godfray et al., 2010). Unlike the average of 9 gallons in the US, the global average for milk production is 2 gallons per day (HolsteinAssociationUSA, 2017).

Improving the output and efficiency globally by just 4.75 oz per cow per day per year is estimated to be sufficient to fill the current demand for milk (Elanco, 2017), and that number must increase with population growth.

The greatest room for growth lies within the poorest producing countries. These tropical climates are too severe for Holsteins, which have been shown to produce poorly and die in the heat and without enough resources to meet their demands (Marshall, 2014). In developing countries, “high output—high input” breeds like Holsteins are generally not successful due to high mortality and morbidity in attempted low-management conditions (Marshall, 2014). Many famers in poor or rural conditions must choose between the “high output—high input” and the

“low output—low input” animals. The low input cattle are usually hardy and survive well in the event of dry years, too much rain or extremely hot temperatures, but production output is limited.

Conversely, the high input cattle are vulnerable and stressed in tropical climates and susceptible to diseases (Marshall, 2014). Crossbred cattle have been explored as a solution to this problem, but production traits or adaptive traits are often compromised. The solution to this problem seems clear: create an animal that is entirely tropically adapted but that has capacity for increased yield in years with good conditions.

Milk yield volume has been shown to depend upon lactose as the major osmole-regulator in synthesis, which in turn is regulated by the alpha-lactalbumin protein concentration. A DNA mutation in the promoter region for the alpha-lactalbumin gene that increases the amount of alpha-lactalbumin expressed has been shown to be correlated with an increase in milk yield in both Holstein and Nelore breeds. The increased amount of alpha-lactalbumin promotes greater

lactose synthesis, which in turn draws in water and drives production. Studies have shown that a short term insulin infusion (decreasing blood glucose availability, therefore limiting a key substrate for lactose synthesis) was shown to decrease milk output with no effect on protein or fat quantity, while the lactose concentration in the milk remained the same, despite decreased volume, supporting the theory that lactose is responsible for controlling milk yield (Holt, 1983;

Rook & Hopwood, 1970). This study also suggests that cows with this mutation are responsive to changes in intake, meaning that in stressful or dry years, tropically adapted cows such as

Nelore with this mutation would still survive well and produce reasonable quantities. But given better nutrition, research suggests that they will out-produce their wildtype counterparts.

In this study, we showed how the CRISPR/Cas9 system could be used in bovine cells for a targeted, single-base substitution as a means to improve genetic yield. Cells with this substitution would harbor a scarless substitution that in theory could be achieved over years through reproductive selection, for example by crossing Nelore and Holstein cattle. If reproduction was used, the crosses would contain alleles from both parents, thereby increasing the milk yield, but also decreasing the tropical adaptivity and giving up some traits essential for production success. By using gene editing, we are able to change one or a few traits, while leaving the majority of the genetics unchanged. The ability to choose exactly what alleles are expressed in offspring provides an invaluable opportunity for livestock production.

Our long term goal for this study is to use Nelore fibroblasts that harbor homozygous A alleles at the (+15) LALBA site for somatic cell nuclear transfer. Further single-cell sequencing must be done to confirm the substitution at this locus, and to check for any off-target mutations created by this data. Milk yield data from these cows compared with their non-edited counterparts can be used to confirm previous research that correlates the (+15) LALBA site with

increased milk yield, and furthermore determine the value that the addition of this trait would have. While the efficacy of this edit is being proved with the cloned heifers, the next step would be to create a homozygous Nelore bull to propagate the dissemination of the desired genotype on a larger scale to make this mutation occur at a high frequency in the Nelore breed, such as it is in

Holstein. Supplementing native genetics with natural mutations such as this one will help us close the food gap and progress into an era with enough food for everyone.

3.4 Figures

Figure 3.1: Map of the CRISPR/Cas9 plasmid. BOVLACTALB indicates the gRNA 20 nucleotide sequences, ligated into the plasmid adjacent to the sequence for the rest of the gRNA scaffold.

Figure 3.2: Genome sequence at BALB locus in the Bos indicus genome, chromosome 5 (NCBI Gene ID:109558698). Red box labeled “gRNA” indicates CRISPR recognition site. Blue line labeled “cut-site” indicates site for double-stranded cleavage by Cas9. The (+15) site is also noted

Figure 3.3: Sequence of the repair template supplied with the CRISPR/Cas9 plasmid. Red box indicates the single base pair substitution at the (+15) site. Repair with this template changes the PAM sequence from TGG to TGA, which no longer allows DNA cleavage.

Figure 3.4: T7 Endonuclease assay on Nelore cells. Lane: 1) ladder, 2) pooled puromycin selected pooled Nelore cells, transfected with plasmid and repair template, 3) pooled GFP sorted Nelore cells, transfected with plasmid and template, 4) non-transfected cells, 5) pooled GFP sorted cells, transfected with plasmid and template, 6) isolated Nelore colony, transfected with CRIPSR plasmid and repair template, 7) non-transfected Nelore cells. Lanes 2, 3, 5, and 6 show faint bands indicating T7 cleavage and sequence variation.

Figure 3.5: Chromatograms from sequences of A) non-transfected Nelore DNA and B) pooled, transfected Nelore DNA. Sequences are consistent and uniform until the CRISPR cut-site (black arrow), and then multiple bases are present at any given position. C) TIDE: Tracking of Indels by DEcomposition (Brinkman et al., 2014) analysis of sequences from A) and B). Analysis indicates that CRIPSR/Cas9 plasmid cleaved DNA at approximately at 26.8% efficiency.

Figure 3.6: BpueI enzyme digest of Nelore cell DNA LALBA PCR product. Lane 1) non- transfected Nelore fibroblast DNA 2) Holstein DNA heterozygous for (+15) mutation, control 3) pooled, transfected Nelore cells 4) isolated transfected Nelore colony. Lane 3 shows cleavage with expected fragment size, while lane 4 shows two bands--on

Chapter 4: General Discussion

Humans have been shaping and selecting the genetics of domestic animals for thousands of years. Breeding domesticated animals to promote certain traits has resulted in the emergence of additional breeds and a great deal of variation in production animals around the world. The speed of genetic selection has increased significantly with the advent of artificial reproductive technologies (ART), which allow us, for example, to take semen from bulls in Canada and mix them with eggs from cows in Brazil, and transfer the resulting embryos to recipients in Kenya.

Breeding is now a global phenomenon.

The entire bovine genome was first sequenced in 2009 (Elsik, Tellam, & Worley, 2009).

Since then, developments in bioinformatics and genetic markers now allow us to select animals as breeding stock based on genetic merit before they are even born. The pace of selection is still accelerating, which is extremely necessary given the demands of the exponential population growth. The next step to continue the necessary advances in production is the incorporation of gene editing.

Initially, gene editing was dependent on low-efficiency homologous recombination, viral transduction, and random integration, and has existed for years. Using this technology, we have gathered knowledge about the role of genes and gene networks, but were limited by low efficiency and little control over where genes were changed. That was, until site specific endonucleases were harnessed as a tool to create targeted breaks in the DNA which can be repaired with strategic methods.

The studies performed in Chapter 2 and Chapter 3 demonstrate the capability of the

CRISPR/Cas9 system to seamlessly introduce single nucleotide changes in the DNA of cattle for

the improvement of genetic merit and production performance. The majority of studies using

CRISPR/Cas9 and other RGENs in livestock have involved knocking out genes, such as myostatin (Crispo et al., 2015; Ni et al., 2014) or prion proteins (Richt et al., 2007), but fewer studies have demonstrated the insertion of genes, and even fewer have demonstrated single nucleotide substitutions in livestock species.

Though we have been changing the livestock genome for years, regulations still prohibit the majority of animals that have been modified using RNA guided endonucleases from being sold in the market. Regulations are necessary to prevent reckless or inhumane engineering of animals. However, in the case of the Aquabounty AquAdvantage salmon, it took more than 20 years for a thoroughly proven safe product to be allowed to reach the markets in the US. A more efficient, comprehensive legal system must be developed for the regulation of transgenic animals. Until then, the genetically repaired Angus cattle and enhanced LALBA Nelore cattle derived from this study cannot be used in production.

Beyond the CRISPR/Cas9 system demonstrated in these experiments, other versions of

CRISPRs exist and are being used today. This commonly used CRISPR/Cas9 from

Streptococcus pyogenes is predicted to be one of thousands similar to it. For example, the Cas9 orthologue found in Staphylococcus aureus is 1kb smaller and has similar targeting efficiency

(Ran et al., 2015), and therefore can more efficiently be packaged in adeno-associated viruses or other methods for gene delivery. Our gene editing toolbox is constantly being expanded and improved. This is especially true as more and more endogenous cellular machinery is understood well enough to harness for our use. For example, the cytidine deaminase and uracil glycosylase, both enzymes found naturally in cells, have now been reconfigured to work together combined with a catalytically inactive Cas9 and gRNA to make specific single base pair edits without even

making double stranded DNA breaks. Termed “base editing” technology, these systems present distinct advantages over the Cas9 used in this research, as the lack of double stranded DNA breakage decreases the possibility of indels, off-target cleavage, and uncontrolled mutation. In future studies, we would like to explore the potential of these base editors in cattle to repair genetic defects or create single base substitutions that we have shown in this study (see appendix). Microinjection of base editor proteins and gRNAs into oocytes and embryos could be preferable to the systems described in Chapters 2 and 3.

It is probable that in the future will have such grasp on the genetic code that we can write our own genes. However the complexity of that skill is beyond any tangible goal at the moment.

Collective efforts are being employed both on the development side of gene editors and on delivery methods. Great progress has allowed for these systems to be employed in vitro and in vivo, for agricultural, biomedical and basic science purposes all around. As this continues to develop, and traits like DD are removed and traits like the (+15) A allele are incorporated, it will be important to look at the big picture for implementing gene editing in agriculture on a larger scale. As discussed, limitations to livestock output are specific to the conditions in each production environment around the world. Studies are continuously being published demonstrating the benefits of different traits being added or knocked out of any given species.

But ideally, to maximize our global agricultural output, we will collectively establish the optimal traits for each livestock species for each global region, and then we will design and produce a series of regionally adapted species with optimized productivity.

References

Adams, H. A., Sonstegard, T. S., VanRaden, P. M., Null, D. J., Van Tassell, C. P., Larkin, D. M., & Lewin, H. A. (2016). Identification of a nonsense mutation in APAF1 that is likely causal for a decrease in reproductive efficiency in Holstein dairy cattle. Journal of Dairy Science, 99(8), 6693-6701. doi:10.3168/jds.2015-10517 Beever, J. (2013). Likely presence of genetic condition in a line of Angus cattle. Bell, J., Beyer, T., & Hill, R. (1976). The kinetic mechansim of bovine milk galactosyltransferase. The role of alpha-lactalbumin. Journal of Biological Chemistry, 251(10), 3003-3013. Blake, J. A., Eppig, J. T., Kadin, J. A., Richardson, J. E., Smith, C. L., & Bult, C. J. (2017). Mouse Genome Database (MGD)-2017: community knowledge resource for the laboratory mouse. Nucleic Acids Research, 45(D1), D723-D729. Bleck, G. T., & Bremel, R. D. (1993a). Correlation of the α-lactalbumin (+ 15) polymorphism to milk production and milk composition of Holsteins. Journal of Dairy Science, 76(8), 2292- 2298. Bleck, G. T., & Bremel, R. D. (1993b). Sequence and single-base polymorphisms of the bovine α- lactalbumin 5'-flanking region. Gene, 126(2), 213-218. doi:https://doi.org/10.1016/0378-1119(93)90369-E Bleck, G. T., White, B. R., Miller, D. J., & Wheeler, M. B. (1998). Production of bovine alpha- lactalbumin in the milk of transgenic pigs. Journal of Animal Science, 76(12), 3072-3078. Brew, K., Vanaman, T. C., & Hill, R. L. (1968). The role of alpha-lactalbumin and the A protein in lactose synthetase: a unique mechanism for the control of a biological reaction. Proceedings of the National Academy of Sciences, 59(2), 491-497. Brinkman, E. K., Chen, T., Amendola, M., & van Steensel, B. (2014). Easy quantitative assessment of genome editing by sequence trace decomposition. Nucleic Acids Research, 42(22), e168-e168. doi:10.1093/nar/gku936 Brown, A., Woods, W. S., & Perez-Pinera, P. (2016). Multiplexed targeted genome engineering using a universal nuclease-assisted vector integration system. ACS synthetic biology, 5(7), 582-588. Campbell, K. H., McWhir, J., Ritchie, W. A., & Wilmut, I. (1996). Sheep cloned by nuclear transfer from a cultured cell line. Nature, 380(6569), 64-66. doi:10.1038/380064a0 Carlson, D. F., Lancto, C. A., Zang, B., Kim, E.-S., Walton, M., Oldeschulte, D., . . . Fahrenkrug, S. C. (2016). Production of hornless dairy cattle from genome-edited cell lines. Nature biotechnology, 34(5), 479-481. Casas, E., & Kehrli, M. E. (2016). A Review of Selected Genes with Known Effects on Performance and Health of Cattle. Frontiers in Veterinary Science, 3, 113. doi:10.3389/fvets.2016.00113 Cong, L., Ran, F. A., Cox, D., Lin, S., Barretto, R., Habib, N., . . . Zhang, F. (2013). Multiplex Genome Engineering Using CRISPR/Cas Systems. Science, 339(6121), 819. Cox, D. B. T., Platt, R. J., & Zhang, F. (2015). Therapeutic Genome Editing: Prospects and Challenges. Nature medicine, 21(2), 121-131. doi:10.1038/nm.3793

Crispo, M., Mulet, A., Tesson, L., Barrera, N., Cuadro, F., dos Santos-Neto, P., . . . Anegón, I. (2015). Efficient generation of myostatin knock-out sheep using CRISPR/Cas9 technology and microinjection into zygotes. PLOS ONE, 10(8), e0136690. Davies, D., Holt, C., & Christie, W. (1983). The composition of milk. Biochemistry of lactation, 71-117. Delhotal, J. D. (2016). CHARACTERIZATION OF NHLRC2 GENE-EDITED MICE: A MODEL FOR BOVINE DEVELOPMENTAL DUPLICATIONS. (Master of Science), University of Illinois at Urbana- Champaign. Denholm, L. J., Martin, L., & Denman, A. (2011). Polymelia (supernumerary limbs) in Angus calves. Retrieved from http://www.flockandherd.net.au/cattle/reader/polymelia.html Denholm, L. J., Parnell, P. F., Marron, B. M., Delhotal, J. D., Moser, D. W., & Beever, J. E. (2016). A comparative review of Developmental Duplications (DD) in Angus cattle and neural tube defects (NTDs) in man and mouse Paper presented at the World Buiatrics Congress, Dublin, Ireland. Dorstyn, L., Kinoshita, M., & Kumar, S. (1998). Caspases in Cell Death. In S. Kumar (Ed.), Apoptosis: Mechanisms and Role in Disease (pp. 1-24). Berlin, Heidelberg: Springer Berlin Heidelberg. Drögemüller, C., & Distl, O. (2006). Genetic analysis of syndactyly in German Holstein cattle. The Veterinary Journal, 171(1), 120-125. Durai, S., Mani, M., Kandavelou, K., Wu, J., Porteus, M. H., & Chandrasegaran, S. (2005). Zinc finger nucleases: custom-designed molecular scissors for genome engineering of plant and mammalian cells. Nucleic Acids Research, 33(18), 5978-5990. doi:10.1093/nar/gki912 Elanco (Producer). (2017). The Protein Gap. Enough Movement. Retrieved from https://www.enoughmovement.com/learn/index.aspx Elsik, C. G., Tellam, R. L., & Worley, K. C. (2009). The Genome Sequence of Taurine Cattle: A Window to Ruminant Biology and Evolution. Science, 324(5926), 522. Esvelt, K. M., & Wang, H. H. (2013). Genome-scale engineering for systems and synthetic biology. Molecular Systems Biology, 9, 641-641. doi:10.1038/msb.2012.66 FAO (Producer). (2016). How to feed the world in 2050. FDA (Producer). (2014, 2017). Animal Cloning. FDA. (2017). AquAdvantage Salmon Fact Sheet. Retrieved from https://www.fda.gov/AnimalVeterinary/DevelopmentApprovalProcess/GeneticEngineer ing/GeneticallyEngineeredAnimals/ucm473238.htm Finger, J. H., Smith, C. M., Hayamizu, T. F., McCright, I. J., Xu, J., Law, M., . . . Blodgett, O. (2017). The mouse Gene Expression Database (GXD): 2017 update. Nucleic Acids Research, 45(D1), D730-D736. Forsberg, C., Phillips, J., Golovan, S., Fan, M., Meidinger, R., Ajakaiye, A., . . . Hacker, R. (2003). The enviropig physiology, performance, and contribution to nutrient management advances in a regulated environment: The leading edge of change in the pork industry. Journal of Animal Science, 81(14_suppl_2), E68-E77.

Gao, Y., Wu, H., Wang, Y., Liu, X., Chen, L., Li, Q., . . . Zhang, Y. (2017). Single Cas9 nickase induced generation of NRAMP1 knockin cattle with reduced off-target effects. Genome Biology, 18(1), 13. doi:10.1186/s13059-016-1144-4 Garneau, J. E., Dupuis, M.-E., Villion, M., Romero, D. A., Barrangou, R., Boyaval, P., . . . Moineau, S. (2010). The CRISPR/Cas bacterial immune system cleaves bacteriophage and plasmid DNA. Nature, 468(7320), 67-71. Godfray, H. C. J., Beddington, J. R., Crute, I. R., Haddad, L., Lawrence, D., Muir, J. F., . . . Toulmin, C. (2010). Food Security: The Challenge of Feeding 9 Billion People. Science, 327(5967), 812-818. doi:10.1126/science.1185383 Greene, N. D. E., & Copp, A. J. (2014). Neural Tube Defects. Annual review of neuroscience, 37, 221-242. doi:10.1146/annurev-neuro-062012-170354 Hai, T., Teng, F., Guo, R., Li, W., & Zhou, Q. (2014). One-step generation of knockout pigs by zygote injection of CRISPR/Cas system. Cell Res, 24(3), 372-375. doi:10.1038/cr.2014.11 Hakem, R., Hakem, A., Duncan, G. S., Henderson, J. T., Woo, M., Soengas, M. S., . . . Mak, T. W. (1998). Differential Requirement for Caspase 9 in Apoptotic Pathways In Vivo. Cell, 94(3), 339-352. doi:https://doi.org/10.1016/S0092-8674(00)81477-4 Hammer, R. E., Pursel, V. G., Rexroad, C. E., Wall, R. J., Bolt, D. J., Ebert, K. M., . . . Brinster, R. L. (1985). Production of transgenic rabbits, sheep and pigs by microinjection. Nature, 315(6021), 680-683. Hansen, P. J. (2014). Current and Future Assisted Reproductive Technologies for Mammalian Farm Animals. In G. C. Lamb & N. DiLorenzo (Eds.), Current and Future Reproductive Technologies and World Food Production (pp. 1-22). New York, NY: Springer New York. Haskell, M. J., Simm, G., & Turner, S. P. (2014). Genetic selection for temperament traits in dairy and beef cattle. Frontiers in Genetics, 5(368). doi:10.3389/fgene.2014.00368 Heigwer, F., Kerr, G., & Boutros, M. (2014). E-CRISP: fast CRISPR target site identification. Nat Meth, 11(2), 122-123. HolsteinAssociationUSA. (2017). Facts about Holstein Cattle. Retrieved from http://www.holsteinusa.com/pdf/fact_sheet_cattle.pdf Holt, C. (1983). Swelling of Golgi vesicles in mammary secretory cells and its relation to the yield and quantitative composition of milk. Journal of theoretical biology, 101(2), 247-261. Horii, T., Arai, Y., Yamazaki, M., Morita, S., Kimura, M., Itoh, M., . . . Hatada, I. (2014). Validation of microinjection methods for generating knockout mice by CRISPR/Cas-mediated genome engineering. Scientific reports, 4, 4513. Jinek, M., Chylinski, K., Fonfara, I., Hauer, M., Doudna, J. A., & Charpentier, E. (2012). A Programmable Dual-RNA–Guided DNA Endonuclease in Adaptive Bacterial Immunity. Science, 337(6096), 816. Joung, J. K., & Sander, J. D. (2013). TALENs: a widely applicable technology for targeted genome editing. Nat Rev Mol Cell Biol, 14(1), 49-55. doi:http://www.nature.com/nrm/journal/v14/n1/suppinfo/nrm3486_S1.html Kelani, A. B., Moumouni, H., Issa, A. W., Younsaa, H., Fokou, H. M. U., Sani, R., . . . Catala, M. (2017). Notomelia and related neural tube defects in a baby born in Niger: case report and literature review. Child's Nervous System, 1-6. doi:10.1007/s00381-017-3337-x Kim, H., & Kim, J.-S. (2014). A guide to genome engineering with programmable nucleases. Nat Rev Genet, 15(5), 321-334. doi:10.1038/nrg3686

Kim, Y. B., Komor, A. C., Levy, J. M., Packer, M. S., Zhao, K. T., & Liu, D. R. (2017). Increasing the genome-targeting scope and precision of base editing with engineered Cas9-cytidine deaminase fusions. Nature biotechnology, 35(4), 371-376. Koch, R. M., Swiger, L. A., Chambers, D., & Gregory, K. E. (1963). Efficiency of Feed Use in Beef Cattle1. Journal of Animal Science, 22(2), 486-494. doi:10.2527/jas1963.222486x Komor, A. C., Kim, Y. B., Packer, M. S., Zuris, J. A., & Liu, D. R. (2016). Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature. Kuhn, N. (1983). The biosynthesis of lactose. Biochemistry of lactation, 159-176. Lander, E. S. (2016). The heroes of CRISPR. Cell, 164(1), 18-28. Lillico, S. G., Proudfoot, C., King, T. J., Tan, W., Zhang, L., Mardjuki, R., . . . Whitelaw, C. B. A. (2016). Mammalian interspecies substitution of immune modulatory alleles by genome editing. Scientific reports, 6, 21645. Long, J., Pan, G., Ifeachor, E., Belshaw, R., & Li, X. (2016). Discovery of Novel Biomarkers for Alzheimer’s Disease from Blood. Disease Markers, 2016, 9. doi:10.1155/2016/4250480 Los, M., Wesselborg, S., & Schulze-Osthoff, K. (1999). The Role of Caspases in Development, Immunity, and Apoptotic Signal Transduction: Lessons from Knockout Mice. Immunity, 10(6), 629-639. doi:https://doi.org/10.1016/S1074-7613(00)80062-X Marshall, K. (2014). Optimizing the use of breed types in developing country livestock production systems: a neglected research area. Journal of Animal Breeding & Genetics, 131(5), 329-340. doi:10.1111/jbg.12080 Martins, L. F., Milazzotto, M. P., Feitosa, W. B., Coutinho, A. R., Simoes, R., Marques, M. G., . . . Visintin, J. A. (2008). Sequence variation of the alpha-lactalbumin gene in Holstein and Nellore cows. Anim Biotechnol, 19(3), 194-198. doi:10.1080/10495390802058939 Massa, V., Savery, D., Ybot-Gonzalez, P., Ferraro, E., Rongvaux, A., Cecconi, F., . . . Copp, A. J. (2009). Apoptosis is not required for mammalian neural tube closure. Proceedings of the National Academy of Sciences, 106(20), 8233-8238. doi:10.1073/pnas.0900333106 Mathupala, S. P., & Sloan, A. E. (2009). An agarose-based cloning-ring anchoring method for isolation of viable cell clones. BioTechniques, 46(4), 305-307. doi:10.2144/000113079 McVey, M., & Lee, S. E. (2008). MMEJ repair of double-strand breaks (director’s cut): deleted sequences and alternative endings. Trends in Genetics, 24(11), 529-538. doi:https://doi.org/10.1016/j.tig.2008.08.007 Meier, P., Finch, A., & Evan, G. (2000). Apoptosis in development. Nature, 407(6805), 796-801. Merlevede, J., Droin, N., Qin, T., Meldi, K., Yoshida, K., Morabito, M., . . . Braun, T. (2016). Mutation allele burden remains unchanged in chronic myelomonocytic leukaemia responding to hypomethylating agents. Nature communications, 7. Mir, S. N., Ullah, O., & Sheikh, R. (2014). Genetic polymorphism of milk protein variants and their association studies with milk yield in Sahiwal cattle. African Journal of Biotechnology, 13(4). Mojica, F., Juez, G., & Rodriguez‐Valera, F. (1993). Transcription at different salinities of Haloferax mediterranei sequences adjacent to partially modified PstI sites. Molecular microbiology, 9(3), 613-621.

Nakade, S., Tsubota, T., Sakane, Y., Kume, S., Sakamoto, N., Obara, M., . . . Sakuma, T. (2014). Microhomology-mediated end-joining-dependent integration of donor DNA in cells and animals using TALENs and CRISPR/Cas9. Nature communications, 5. Neupane, M., Moss, K., Avila, F., Raudsepp, T., Marron, B., Beever, J., . . . Neibergs, H. (2017). Case Study: Polymelia in a Holstein calf. The Professional Animal Scientist, 33(3), 378- 386. Newton, J. (2016). Price Transmission in Global Dairy Markets. International Food and Agribusiness Management Review, 19(B). Ni, W., Qiao, J., Hu, S., Zhao, X., Regouski, M., Yang, M., . . . Chen, C. (2014). Efficient Gene Knockout in Goats Using CRISPR/Cas9 System. PLOS ONE, 9(9), e106718. doi:10.1371/journal.pone.0106718 Nicholas, F. (2017). Online Mendelian Inheritance in Animals (OMIA). In U. o. Sydney (Ed.). Nishida, K., Arazoe, T., Yachie, N., Banno, S., Kakimoto, M., Tabata, M., . . . Hara, K. Y. (2016). Targeted nucleotide editing using hybrid prokaryotic and vertebrate adaptive immune systems. Science, 353(6305), aaf8729. Pardo, B., Gómez-González, B., & Aguilera, A. (2009). DNA Repair in Mammalian Cells. Cellular and Molecular Life Sciences, 66(6), 1039-1056. doi:10.1007/s00018-009-8740-3 Ran, F. A., Cong, L., Yan, W. X., Scott, D. A., Gootenberg, J. S., Kriz, A. J., . . . Zhang, F. (2015). In vivo genome editing using Staphylococcus aureus Cas9. Nature, 520(7546), 186-191. Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott, D. A., & Zhang, F. (2013). Genome engineering using the CRISPR-Cas9 system. Nat. Protocols, 8(11), 2281-2308. Richt, J. A., Kasinathan, P., Hamir, A. N., Castilla, J., Sathiyaseelan, T., Vargas, F., . . . Kuroiwa, Y. (2007). Production of cattle lacking prion protein. Nat Biotech, 25(1), 132-138. doi:10.1038/nbt1271 Rook, J., & Hopwood, J. (1970). The effects of intravenous infusions of insulin and of sodium succinate on milk secretion in the goat. Journal of Dairy Research, 37(02), 193-198. Shimatani, Z., Kashojiya, S., Takayama, M., Terada, R., Arazoe, T., Ishii, H., . . . Miura, K. (2017). Targeted base editing in rice and tomato using a CRISPR-Cas9 cytidine deaminase fusion. Nature biotechnology. Simmons, J. (2011). Making safe, affordable and abundant food a global reality. Su, F., Wang, Y., Liu, G., Ru, K., Liu, X., Yu, Y., . . . Guo, Z. (2016). Generation of transgenic cattle expressing human β‐defensin 3 as an approach to reducing susceptibility to Mycobacterium bovis infection. FEBS Journal. Suzuki, K., Tsunekawa, Y., Hernandez-Benitez, R., Wu, J., Zhu, J., Kim, E. J., . . . Li, Z. (2016). In vivo genome editing via CRISPR/Cas9 mediated homology-independent targeted integration. Nature, 540(7631), 144-149. Tang, L., González, R., & Dobrinski, I. (2015). Germline modification of domestic animals. Animal reproduction / Colegio Brasileiro de Reproducao Animal, 12(1), 93-104. Van Eenennaam, A. L. (2014). Managing Genetic Defects. In eXtension (Ed.). ebeef.org. Van Eenennaam, A. L. (2017). Genetic modification of food animals. Current Opinion in Biotechnology, 44, 27-34. doi:https://doi.org/10.1016/j.copbio.2016.10.007 Varfolomeev, E. E., Schuchmann, M., Luria, V., Chiannilkulchai, N., Beckmann, J. S., Mett, I. L., . . . Kollet, O. (1998). Targeted disruption of the mouse Caspase 8 gene ablates cell death

induction by the TNF receptors, Fas/Apo1, and DR3 and is lethal prenatally. Immunity, 9(2), 267-276. Wall, R. J., Powell, A. M., Paape, M. J., Kerr, D. E., Bannerman, D. D., Pursel, V. G., . . . Hawk, H. W. (2005). Genetically enhanced cows resist intramammary Staphylococcus aureus infection. Nat Biotech, 23(4), 445-451. Walstra, P. (1999). Dairy technology: principles of milk properties and processes: CRC Press. Whitworth, K. M., Rowland, R. R. R., Ewen, C. L., Trible, B. R., Kerrigan, M. A., Cino-Ozuna, A. G., . . . Prather, R. S. (2016). Gene-edited pigs are protected from porcine reproductive and respiratory syndrome virus. Nat Biotech, 34(1), 20-22. Wilmut, I., Schnieke, A. E., McWhir, J., Kind, A. J., & Campbell, K. H. (1997). Viable offspring derived from fetal and adult mammalian cells. Nature, 385(6619), 810-813. doi:10.1038/385810a0 Woo, M., Hakem, R., Soengas, M. S., Duncan, G. S., Shahinian, A., Kägi, D., . . . Kaufman, S. A. (1998). Essential contribution of caspase 3/CPP32 to apoptosis and its associated nuclear changes. Genes & development, 12(6), 806-819. Wu, H., Wang, Y., Zhang, Y., Yang, M., Lv, J., Liu, J., & Zhang, Y. (2015). TALE nickase-mediated SP110 knockin endows cattle with increased resistance to tuberculosis. Proceedings of the National Academy of Sciences, 112(13), E1530-E1539. doi:10.1073/pnas.1421587112 Yamaguchi, Y., Shinotsuka, N., Nonomura, K., Takemoto, K., Kuida, K., Yosida, H., & Miura, M. (2011). Live imaging of apoptosis in a novel transgenic mouse highlights its role in neural tube closure. The Journal of Cell Biology, 195(6), 1047-1060. doi:10.1083/jcb.201104057 Yoneda, K., Moritomo, Y., Takami, M., Hirata, S., Kikukawa, Y., & Kunieda, T. (1999). Localization of a locus responsible for the bovine chondrodysplastic dwarfism (bcd) on chromosome 6. Mammalian genome, 10(6), 597-600. Yoshida, H., Kong, Y.-Y., Yoshida, R., Elia, A. J., Hakem, A., Hakem, R., . . . Mak, T. W. (1998). Apaf1 Is Required for Mitochondrial Pathways of Apoptosis and Brain Development. Cell, 94(6), 739-750. doi:https://doi.org/10.1016/S0092-8674(00)81733-X Zong, Y., Wang, Y., Li, C., Zhang, R., Chen, K., Ran, Y., . . . Gao, C. (2017). Precise base editing in rice, wheat and maize with a Cas9-cytidine deaminase fusion. Nature biotechnology, 35(5), 438-440.

Appendix A: Additional Figures

Figure A.1: Transfection optimization of dermal fibroblasts using GFP plasmid and Fugene HD transfection reagent.

Figure A.2: Full list of sequences from edited homozygous DD cells (continued from Figure 2.7)

Figure A.3: RFLP using enzyme BpueI of PCR products extracted from single-cell Nelore colonies transfected with the CRISPR/Cas9 plasmid and sorted. Further examination showed that this CRISPR did not cut at the desired locus, and as a result the gRNA was re-ligated into plasmid.

Appendix B: Protocols for Transfections and Analysis

B.1 Transfection of Bovine Fetal Fibroblasts with FuGENE HD

1) Seed cells 24 hrs before transfection at 75,000 cells/well in 6 well plate (15,000 in a 24 well plate) in normal growth medium. 2) 1 hr before transfection change media to growth media without antibiotics. 3) Using a 3 µl FuGENE: 1ug DNA ratio (optimized based on cell type), calculate the amount of each reagent needed.

Example: 2 wells

Opti-MEMTM 280µl

FuGENE HD 12µl

Plasmid DNA 4 µg

4) Add FuGENE HD to tube with Opti-MEMTM reduced serum medium. 5) Add plasmid DNA to Opti-MEMTM and FuGENE tube. 6) Vortex gently and centrifuge briefly. Incubate at room temperature for 15 min. 7) Add 150µl of the tube to each well. 8) Return plate to the growth incubator for 24 hrs. 9) After 24 hrs, remove transfection media and replace with normal growth media (including antibiotics). 10) Examine UV fluorescence under microscope to estimate transfection efficiency if plasmid has GFP indicator.

B.2 Preparation of Transfected Cells for Fluorescence Activated Cell Sorting

1) 24-48 hours after transfection with a fluorescent protein, estimate transfection efficiency with fluorescent microscope. 2) Trypsinize each well according to cell culture protocols as normal: rinse with PBS, add trypsin, allow for cells to detach, then add equal part DMEM. 3) Combine all wells transfected with the same plasmid, centrifuge at 1000 x g for 5 minutes. Resuspend pellet in normal DMEM (with antibiotic) at 1 million cells/mL. 4) Determine whether cells will be sorted into 6-well, 24-well, 96-well plates or a tube. Fill the desired plates with an appropriate amount of media. 5) Bring samples to FACS facility. Be sure to bring a non-transfected control sample to establish parameters for each cell type. 6) Transfer samples to tubes compatible with FACS machine. Add propidium iodide (2µl/ml media) to sample to stain dead cells. *Be sure that transfection reagent has been removed for at least 24 hrs to prevent propidium iodide from staining live cells. Do not use propidium iodide if transfection reagent has not been removed for over 24 hrs. 7) Sort samples into plates. For 96 well plates, sort 3-5 cells per well and centrifuge plates briefly to help cells attach to bottom. For sensitive cells, 25% conditioned media can be used to improve survival. 8) Allow cells to grow to confluency before passaging to larger well or extracting DNA for further analysis.

B.3 Quick DNA extraction from Cells

1) Harvest cells from plates and centrifuge at 1000 xG for 5 min to pellet. Resuspend in PBS to wash, transfer to 0.5ml PCR tube, and then centrifuge again to pellet. 2) Add approximately 100 µl of QuickExtractTM per million cells (lower volume can be used for fewer cells). 3) Vortex to resuspend pellet and centrifuge briefly to collect all material on bottom of tube. 4) Using thermal cycler, incubate samples to digest at 65C for 2 hrs, then stop reaction by denaturing at 95C for 10 minutes. 5) Store DNA at -20C until further analysis

B.4 PCR using Taq JumpStart Readymix

1) Prepare master mix of components: Taq JumpStart Readymix 12.5 µl Forward Primer (10 uM concentration) 1 µl Reverse Primer (10uM concentration) 1 µl Water 8.5 µl 2) Add 23 µl of master mix to 0.5 mL PCR tubes. 3) Add 2 µl of DNA to each tube (approximately 50-500ng depending on target). Vortex gently and centrifuge briefly to collect all material at bottom of tube. 4) Load samples into thermal cycler, using heated lid. PCR thermal cycler conditions must be optimized based on length of fragment and primers. Initial parameters: 5 minutes at 95°C 30 seconds at 94°C, 30 seconds at 56°C and 30 seconds at 72°C for 30 cycles 15 minutes at 72°C 5 min at 10°C Hold at 4°C 5) During reaction, pour 2% agarose gel for electrophoresis and add appropriate comb for number of samples. 6) After reaction, add 1:6 dilution of loading dye per tube (5ul per 25ul sample). Pipet gently and load 5-10 µl of sample into each well. Always include ladder and positive control. 7) Gel electrophoresis at 85-90V for approximately 20-40 minutes, depending on size of amplicon. 8) Use UV light (GelReader) to analyze DNA.