THE BIOLOGICAL MECHANISM BEHIND EARLY AND LATE APPLE SPORTS By Alexander J. Engelsma A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Horticulture – Master of Science 2024 ABSTRACT Somatic mutations in apple commonly develop into viable bud sports that can be propagated clonally. When the apple bud sport has a desirable attribute such as improved color, size, shape, flavor, firmness, sweetness, or harvest timing, it has potential to be introduced as a new cultivar that growers utilize, and consumers enjoy. The genetic mutations and related mechanisms associated with early or delayed maturation (respectively resulting in early or late harvest date) in apple sports are not known despite their value to the industry. By acquiring knowledge about genetic mutations affecting harvest date and their respective molecular mechanisms, breeders can identify markers to conduct more informed crosses to select for early or late maturing apple lines. In this study, the late maturing ‘Gala’ sport ‘Autumn Gala’, the early maturing ‘Fuji’ sport ‘September Wonder Fuji’, and the early maturing ‘Cripps Pink’ (‘Pink Lady®’) sport ‘Maslin Cripps Pink’, were compared to the controls for each cultivar (i.e., those possessing standard harvest times). We found that in each comparison, fruit growth rate of the early variant was significantly greater early in fruit development, during the cell division phase. The early emergence of phenotypic differences in growth rate between the bud sport and the control lines suggests the physiological processes leading to an early or late harvest date may also emerge very early in fruit development. If so, the early or delayed maturation date is very likely not strictly a function of ripening-related processes, but rather is derived from a season-long shift in metabolic activity. Genomic analyses were also done to identify genetic differences between early and late apple sports. Collectively, hundreds of genetic variants were identified. Our phenological studies reduced the developmental window for these transcriptomic investigations. Copyright by ALEXANDER J. ENGELSMA 2024 To Grace and Raine iv ACKNOWLEDGEMENTS On my wedding day I thought I would acknowledge all the people that led my wife Grace and I to that moment. Likewise, I originally planned to here acknowledge all the shoulders upon whom I relied through this unique part of my life in graduate school. After just a few moments in thought, I knew I would fail to mention everyone. As in life, it is impossible to sufficiently thank everyone or repay the debt of mentorship and love. Here I acknowledge a few select individuals and their unique contribution to my mindset and worldview that tremendously assisted me in graduate school. To my wife, Grace, who has loved and supported me through the past ~3 years of marriage. She gave me so much inspiration from her work ethic, love for family, love for our daughter Raine, and insurmountable intelligence. To my parents, both mine and Grace’s. My Dad taught me many things, due likely to his appreciation for learning. Specifically, he taught me to work one piece at a time and persevere. Ma taught me contentment and joy in all things. My father- and mother-in-law taught me to think critically and conscientiously, respectively. To my Uncle Jim, who taught me a great deal of life lessons while working on his apple farm. He instilled in me how to balance self-worth and self-sacrifice, wealth and honor. Hard work on the farm taught me that I could do anything if I set my mind to it. He was the first to recommend that I pursue graduate school. To Courtney and Randy, who inspire me in so many ways to be the best researcher I can be. To Jesus Christ, my savior. “Nothing in my hands I bring, simply to the cross I cling.” v TABLE OF CONTENTS LIST OF ABBREVIATIONS ...................................................................................................... viii CHAPTER 1. Literature Review ..................................................................................................... 1 CHAPTER 2. Apple bud sport comparisons to progenitor/standard harvest cultivar of fruit growth suggest physiological predetermination of maturation date early in fruit development ... 16 CHAPTER 3. Genetic underpinnings of early and late maturing apple bud sports ...................... 52 CHAPTER 4. Conclusions and Future Direction .......................................................................... 85 LITERATURE CITED ................................................................................................................. 89 APPENDIX ................................................................................................................................... 96 vi LIST OF ABBREVIATIONS A: photosynthetic rate (µmol CO2 m-2 s-1) Apple bud sport: cultivars originating not from cross pollination and seed, but from mutated meristematic tissue in the bud of its parent progenitor Cultivar: specific genotype (e.g., 'Gale Gala', 'Imperial Gala', 'Royal Red Honeycrisp', 'Premier Honeycrisp', 'September Wonder Fuji', 'Yataka Fuji', 'Aztec Fuji') DAFB: days after full bloom DEGS: differentially expressed genes Development: the span of growth, differentiation of tissue, and physiological behavior from fertilization of apple flower until harvest GDH: growing degree hours LG: linkage groups Maturation: stage when the fruit is mature and ready to ripen, and envelops starch degradation, the beginning of pre-climacteric and climacteric ripening postharvest Maturity: indicator of harvest date Ripening: ethylene induced developmental stage when the fruit prepares itself for the consumer Variety: a collection of specific genotypes (cultivars) marketed uniformly as one name to the consumer (e.g. 'Gala', 'Honeycrisp', 'Fuji', 'Red Delicious') vii CHAPTER 1. Literature Review 1.1. INTRODUCTION New varieties and cultivars of apples often originate through genetic mutations in meristematic tissues. These natural phenomena result in something called a bud sport, which can be vegetatively propagated. Apple bud sports may develop apple fruit with a phenotype that differs from that of the progenitor. Phenotypic differences in apple fruit may become apparent at any time during development and can impact apple fruit quality (e.g., skin color, flesh color, development rate, maturation date). Developing early- and late-harvesting apple sports is important not only for economic value, saving grower return on investment and dominating an earlier fresh apple market, but also feeding the world. Consumer preference has required apple growers to plant and produce fruit with desirable visual and eating characteristics while maintaining marketability. As a result, apple growers continually plant new cultivars to remain competitive in their markets. Therefore, when a new apple bud sport develops and is commercially available, growers often have the incentive to grow that unique cultivar if it fits into their marketing program. It has been shown in a comprehensive study that consumer preference is primarily driven by firmness in the apple fruit and fruit firmness interplay with soluble solid content (SSC) and titratable acidity (TA) (Harker et al., 2008). Growers, however, prioritize color and storability over flavor and texture. Successful cultivars generally contain several of these qualities. Successful apple bud sports typically retain all or most of the traits of the progenitor with additional qualities, such as enhanced color, extended shelf life (may imply higher fruit firmness) and altered maturation. On an apple grower’s orchard, there are many things to consider throughout the growing season, such as cultural practices of protection, nutrition, fruit quality, the harvest process, and its 1 timeline. The timeline of harvest is critical for growers. If a grower has a primarily ‘Gala’ and ‘Honeycrisp’ farm, those two varieties have a standard harvest window at nearly the same time, the first/second week of September in the Grand Rapids, Michigan area (Shane & Lavely, 2023). It is difficult for a grower to hire enough help or harvest fruit fast enough in this timeframe depending on the acreage, so spreading out the harvest window is the best option to obtain full yield potential. Currently, growers utilize plant growth regulators (PGRs) to delay ripening and maintain apples on the tree for a longer period (AgroFresh, n.d.; Argenta et al., 2022; Bramlage et al., 1980). In the apple industry to date, the use of PGR’s is economically and logistically one of the best strategies to broaden this harvest window, as most growers in the Michigan area utilize these products. Although these products are effective, their cost is substantial, amounting to several hundred dollars per acre (J. Engelsma & E. Ott, personal communication, 2024) with apple prices at 751 bushels per acre amounting to about $14.40 per bushel, $10,816 gross per acre. Investigations into the physiological and genetic background of apple development can lead to discoveries that help growers improve control over the harvest window and protect their investments by reducing overall harvest costs. Future discoveries in the genetic background of apple development and maturation may provide more tools for growers to manipulate their own harvest. Apples that are late harvesting may be impacted by freezing temperatures, which can significantly injure the crop (Chassagne-Berces et al., 2009). This is a primary reason why growers in the Michigan area planted early harvesting sport ‘Maslin Cripps Pink’ (~2 weeks early) instead of its progenitor ‘Cripps Pink’ which harvests in the first week of November, a week with high probability of freezing temperatures in Michigan (Isard & Schaetz, 1998). In addition, Michigan growers must focus on storage, shipping and managing postharvest issues in 2 the late fall. Early harvesting cultivars also present a positive trait of shorter growing season, thus less cost, sprays, water resources, and less cumulative risk due to pest or weather-related issues. All these positive traits make early harvesting cultivars attractive to grow for reasons other than return on investment. Further, the greater control over harvest window may be of benefit in a warming climate. Positive cropping traits that combat our changing climate would benefit the entire apple industry. 1.2. APPLE BUD SPORTS 1.2.1. Background A ‘bud sport’ is a lateral shoot, inflorescence or single flower/fruit with a visibly different phenotype from the rest of the plant; it develops in response to a mutation within the somatic cells of the meristematic tissue (Foster & Aranzana, 2018). These genetic events are found frequently in apple (Malus  domestica), in some varieties more commonly than in others. There are bud sports for select traits such as early harvesting sports (e.g., ‘Premier Honeycrisp’, ‘Yataka Fuji’, ‘Wildfire Gala’), late harvesting sports (e.g., ‘Autumn Gala’), and enhanced color sports (e.g., ‘Firestorm Honeycrisp’, ‘Royal Red Honeycrisp’, ‘Aztec Fuji’, ‘Gale Gala’, ‘RubyMac’). Successful bud sports most often contain all desirable characteristics of the progenitor with one, rarely two or more, additional desirable traits exceeding the quality of the progenitor. 1.2.2. History The first known instance of an apple bud sport observation was made by Peter Collinson in 1741 (Linné et al., 1821). He wrote to a friend of his, mentioning he had found russet apples growing on two thirds of what originally was a green apple tree. Charles Darwin observed similar spontaneous mutations in many plant species and mentions them in his book, The Variation of 3 Plants and Animals Under Domestication (Darwin, 1868). Apple bud sports grew in popularity during the early 1900’s, according to a review by Shamel and Pomeroy (1936). They reported numerous findings in which hundreds of apple bud sports were found. After news of growers patenting their bud sports, there was a race to find anomalies in the orchard. Most valuable apple bud sports were red sports, having more red color than the progenitor (Zotta, 2015). The value of red apple sports was driven by consumer preference of a red apple. There were also many cases in which apple bud sports differed in maturity. Early maturing bud sports were often preferred because of their rapid entry into the markets. Shamel and Pomeroy (1936) informed growers how to look for bud sports and discussed the importance and value of bud mutations in horticultural crops. 1.3. APPLE FRUIT DEVELOPMENT 1.3.1. Pre-harvest Fruit Development Apple fruit development is a complex process involving several stages that take place from floral bud break until the apples are harvested. Before fruit fertilization, flower bud induction and initiation take place during the previous year of fruit development. After a period of dormancy during the winter, flower buds resume growth (break) in the early spring of the following season, progressing through the following stages: silver tip, green tip, quarter-inch green, half-inch green, tight cluster, (first) pink, full pink, king bloom, full bloom, petal fall, 8-mm fruitlet, and 10-mm fruitlet, as described by Michigan State University Extension (MSUE, 2014). Apple fruit development is commonly divided into 3 stages, cell division, cell expansion, and maturation. Following the latter stage, the apple ripens, and this can happen on or off the tree. Once fruit development initiates after fertilization, early fruitlet cortex growth is primarily due only to cell division for about the first 7-10 days after bloom, although it continues until about 4 35-40 days after bloom (Goffinet et al., 1995). Cell division and cell expansion overlap a few weeks after bloom, and cell expansion, primarily of the cortex, continues until harvest (A. N. Lakso, 2011). Cell division is very important to final fruit set and fruit growth, which is highly regulated by temperature and photosynthetically active solar radiation (A. N. Lakso, 2022). Goffinet et al. (1995) found that cell division during the first 3 weeks after bloom controls the maximum final fruit size. Apple fruit growth is highly sensitive to temperature in the first 42 days (Bergh, 1990). Effects of solar radiation on fruitlet abscission are primarily controlled by the demand of carbon by the fruit and that is markedly influenced by temperature (A. N. Lakso et al., 2001). At lower temperatures, low solar radiation isn’t as inhibitory to fruit growth as it is at high temperatures (A. N. Lakso, 2022). Fruit may thin easily during high rates of cell division due to strong carbon demand, and under high temperatures and cloudy conditions thinning may be further promoted. Cell expansion begins roughly 7-10 days after bloom (Goffinet et al., 1995) and continues until harvest. The phase of cell expansion overlaps both cell division and maturation/ripening phases. Expansion of cells is less correlated with whole fruit growth, as Goffinet et al. (1995) found that fruit size was positively correlated with cell number, not cell size or proportion of intercellular space (Goffinet et al., 1995). After cell division ceases at ~40 days after bloom, cell expansion continues until harvest. 1.3.2. Ripening Ripening is another complex process which involves many different reactions and substrates. Ripening in apple is driven by the gaseous plant hormone ethylene. In postharvest physiology, ethylene production displays itself through two different systems during fruit ripening (Biale, 1964; Pratt & Goeschl, 1969). Apples are climacteric, meaning that apples continue to ripen after harvest, involving a burst in respiration. The ethylene production system (system II) associated 5 with ripening is autocatalytic, in that the reaction product, ethylene, is perceived and stimulates more ethylene production. The other system of ethylene production is in all fruits, climacteric and non-climacteric. This latter system (system I) does not result in a burst of ethylene, but a steady perception and limits ethylene production to small quantities because it is self-inhibitory. As ethylene accumulates within the apple fruit via system II, the apple rapidly ripens, and its qualities (e.g. firmness, color, flavor, starch degradation, simple sugar accumulation) develop for a desirable eating experience. Another popular method for tracking maturation and ripening of apples is by assessing starch degradation, which occurs before higher ethylene production (system II) (Blanpied & Silsby, 1992; Brookfield et al., 1997; Thammawong & Arakawa, 2007), by staining the inner cortex and core with an iodine solution. This solution stains the starches in the fruit and, as the fruit ripens, the starch declines and converts to simple sugars that cannot be stained by the iodine solution. 1.3.3. Ripening Comparisons The nature of behavioral differences in ripening between apple varieties is well-studied (Giné- Bordonaba et al., 2019; Gussman et al., 1993; Song et al., 1997; Tong et al., 1999). The sensitivity of apple fruit to ethylene has been analyzed across varieties with differing harvest dates (Singh et al., 2017). Singh et al. (2017) found that early-maturing apple variety ‘Anna’, developed in Israel, produces higher amounts of ethylene, and has higher respiration rates in comparison to later harvesting varieties ‘Galaxy’ and ‘Golden Delicious’. Not only are the levels of ethylene and respiration higher, but the ‘Anna’ cultivar exhibits properties of system II ethylene production (auto stimulatory) throughout fruit development (Singh et al., 2017). Cultivar ‘Anna’ responded to exogenous ethylene treatments during the early stages of fruit development, while the late harvesting cultivars did not. Loss of postharvest quality and early 6 maturation in ‘Anna’ may be a result of higher ethylene production earlier in fruit development. Comparisons of early and late sports of the same apple variety have also been made (Dong et al., 2011; Iglesias et al., 2012; Kim et al., 2023; Wang et al., 2009). In the case of Wang et al. (2009), they compared mutant cultivar ‘Hirosaki Fuji’ to the progenitor, ‘Fuji’. In their study, they find different expression levels of several ethylene biosynthesis and perception genes, and a heat shock protein coding gene. Although the data is unpublished, Wang et al. (2009) mention that fruit diameter data indicate a faster rate of growth in the early sport as opposed to the standard harvesting ‘Fuji’. They describe the implication of the data by mentioning that the mutation responsible for the early maturation in ‘Hirosaki Fuji’ is unrelated to the difference in expression of genes involved in ripening processes (Wang et al., 2009). 1.4. PHOTOSYNTHESIS AS A DRIVER OF FRUIT DEVELOPMENTAL RATE 1.4.1. Background Photosynthesis has been extensively studied in apple, primarily in relation to stress or field environment of apple trees (Bhusal et al., 2019; DeJong, 2022; Grappadelli et al., 1994; A. N. Lakso, 1983; Schneider & Childers, 1941; X. Sun et al., 2018; Tartachnyk & Blanke, 2004; Z. Wang et al., 2018). Much of the past and current research on photosynthesis in apple trees relates to response during different light conditions, seasonal conditions, stress or environmental conditions, and low or high nutrition conditions. 1.4.2. Photosynthesis Methodology in Apple The method of measuring photosynthesis in apple trees is important to review and understand due to the complexity of the photosynthetic mechanism and the biology of an entire tree organism. Infrared gas analyzers have been utilized for quite some time (Fastie & Pfund, 1947). In apple tree fruit photosynthesis, primarily two different infrared gas analyzer systems are used: 7 open and closed systems (Flore & Lakso, 1989). In an open system, air is continuously flowing through the chamber, allowing steady-state to be reached between the chamber headspace and the leaf. In a closed chamber system, the leaf is put into the chamber and the measurements are immediately read as the CO2 is taken up by the leaf, creating a CO2 deficit that is measured by the instrument. Both systems are viable for measuring leaf photosynthesis, but the open chamber is preferred due to the ability to control temperature, humidity, and CO2 concentrations (Trimble, 2020). 1.4.3. Units A very important element of measuring photosynthesis to consider is what is being measured. Interpretation of photosynthesis data relies on what plant part is used in the photosynthetic rate measurements (Flore & Lakso, 1989). In peach, Kappel and Flore (1983) measured photosynthetic assimilation (A) rate in 4 ways: A per leaf area, A per whole leaf, A per mg chlorophyll, and A per unit dry weight (Kappel & Flore, 1983). Flore and Lakso recalculated Kappel and Flore’s work to update units (mass to mol). They found that, based upon light levels described as photosynthetic photon flux (PPF), measurements based on dry weight decreased with an increase in PPF, while A expressed per unit chlorophyll increased with increasing PPF (Fig. 1). Thus, research questions for a project may heavily depend upon which measure of A is considered. Measuring A of CO2 as µmol m-2 s-1 is an appropriate method to capture total carbon assimilated by leaf area for an entire day. 8 Figure 1. Relative photosynthetic carbon assimilation of peach leaves grown under different levels of PPF, expressed as A per leaf area (LA-1), A per whole leaf (leaf-1), A per mg chlorophyll (chl-1) and A per unit dry wt-1(d wt-1). (Recalculated from Kappel and Flore 1983). Figure taken from Lakso and Flore, 1989. 9 1.4.3. Diurnal and Seasonal Photosynthetic Activity Diurnal measurements are an important method of analysis in determining peak A, decreases in A, and A throughout a day at specific times. There is added benefit by capturing diurnal measurements due to potential trends throughout the day that the apple tree is exhibiting, not detectable by just one measurement. Diurnal measurements also allow one to integrate total carbon fixed for the day. These trends of carbon fixation may vary throughout the day due to a few reasons, one reason being shifts in stomatal conductance. Stomatal conductance, a critical factor in photosynthesis, is the measurement of the ease of gas exchange through the stomatal aperture. Interplay between stomatal conductance, temperature, and humidity can play a large role in an increase or decrease of A (Flore & Lakso, 1989). Stomatal conductance and net photosynthesis are very linearly correlated (A. N. Lakso, 1979). Thus, increases or decreases in photosynthetic rate throughout the day may heavily rely on the environmental conditions of the day, such as vapor pressure gradient and temperature, and their effect on stomatal conductance. Seasonal changes result in a gradual decline in net photosynthesis approaching harvest, the highest rate of photosynthesis during the season averaging a net CO2 flux of ~15-22 µmol m-2 s-1 (Flore & Lakso, 1989). A comprehensive study performed by Fujii and Kennedy (1984) revealed two periods of time during the growing season that photosynthetic rates differed from each other according to fruit load (Fujii & Kennedy, 1984). Photosynthetic rates first differed during bloom, exhibiting a 25% increase in photosynthetic rate on flowering shoots compared to vegetative shoots. Fujii and Kennedy (1984) found that at a later period in the growing season, from July to September, CO2 assimilation rates were higher in fruiting trees than nonfruiting trees. From the findings of this study, it is important to note the importance of crop load on apple leaf or tree 10 photosynthesis because the crop load amount on the tree may determine and/or influence net photosynthetic rate. There is little literature comparing one variety to another regarding carbon assimilation, stomatal conductance, and photosystem II efficiency. Most, if not all research in apple photosynthesis has revealed environmental, exogenous hormone application, nutrient level, and water effects on photosynthetic rate, photorespiration, photochemistry, dark respiration, and stomatal conductance. Furthermore, to the author’s knowledge, there is no literature on photosynthesis analyses between apple cultivars within the same variety. Additionally, no known comparison has been conceived between early/late maturing sports and their progenitor/standard harvest time cultivar. Further knowledge and understanding of carbon assimilation rates amongst cultivars of the same apple variety may provide answers as to why there is a stark difference in maturity time between a progenitor and its early or late maturing bud sport. If development before harvest is compressed or extended, net photosynthesis data may provide an explanation for that advanced or delayed development. Photosynthesis may limit fruit development, or on the contrary, fruit growth may regulate net photosynthetic rates. A lower photosynthetic rate may delay fruit development or, conversely, lower rates of fruit development may result in lower demand on the leaves, leading to a decline in photosynthetic rate. 1.5. APPLE GENOMICS AND TRANSCRIPTOMICS 1.5.1. Background Only a handful of select widely cultivated varieties of apple (Malus domestica) have been comprehensively sequenced (Daccord et al., 2017; Khan et al., 2022; X. Sun et al., 2020); however this number is likely to grow exponentially as prices decline and sequencing technologies advance. Primary reasons for sequencing the ‘Honeycrisp’ genome were because of 11 its high value to the apple industry and preference of the consumer, in addition to the major problems it presents after harvest (Khan et al., 2022) such as internal browning (Contreras et al., 2014) and bitter pit (Griffith, 2022). Sequencing apple cultivars has the potential to contribute to an expanded understanding of genetic heritage, evolution, predisposition to disease and physiological disorders, as well as breeding new cultivars for enhanced quality, resistance to disease, robust photosynthesis, enhanced storability and, in the case of the project discussed hereafter, harvest date manipulation. 1.5.2. Gene Mutation in Apple Fruit Qualities There have been several studies regarding mutations causing phenotypic change in apple quality (ethylene production, color, volatile production, and overall ripening behavior) (Cho et al., 2020; Dong et al., 2011; Kim et al., 2023; H. Sun et al., 2023; Telias et al., 2011). Few studies (Ban et al., 2022; Kim et al., 2023; Wei et al., 2020) have analyzed mutations resulting in altered harvest dates in apple or other fruit crops, despite the value that the background knowledge of genetic predetermination of harvest date may bring. In the case of the red sport ‘RubyMac’ they identified altered candidate genes that may play a significant role in fruit coloration (Sun et al., 2023). Comparisons between bud sports of the same variety have been investigated with regard to several fruit qualities including ethylene (Dong et al., 2011), color and anthocyanin content (Cho et al., 2020; Telias et al., 2011), volatile compounds (Iglesias et al., 2012) and overall fruit maturation, ripening, and ripening gene expression (Kim et al., 2023). Dong et al. (2011) found that early maturing ‘Fuji’ cultivar ‘Beni Shogun’ displayed earlier and higher expression of key genes related to ethylene synthesis, signaling, and transduction than ‘Fuji’. Cho et al. (2020) found genes whose expression was associated with the skin color of ‘Fuji’ cultivars at mature stages, as well as higher expression of MdMYB10 and MdGST genes related to anthocyanin 12 production in ‘Beni Shogun’ cultivar compared to ‘Fuji’. Iglesias et al. (2012) found volatile compound concentration and profile differences among ‘Fuji’ cultivars. They found the lowest concentration in ‘Aztec Fuji’. They also show correlations through a full-data principal component analysis between consumer preference and volatile profiles in the different ‘Fuji’ cultivars. Study of mechanisms behind early and late bud sports is not limited only to apple. Other species such as grape, peach, and pear are also being evaluated for their maturity differences (X. Liu et al., 2014; Y. Liu et al., 2007; Wei et al., 2020; Wu et al., 2015) 1.5.3. Gene Expression in Grape Cultivars Differing in Harvest Date In grape, Wei et al. (2020) discover candidate genes, differentially expressed between an early- ripening cultivar and its progenitor (Wei et al., 2020). Some of the differentially expressed genes (DEGS) are thought to be involved in berry ripening, implying possible mechanisms for the early ripening phenotype of the grape bud mutant. Another study compared two bud sports to their parent in a transcriptomic analysis with the intention of confirming the two bud sports as reliable cultivars for cultivation (Wu et al., 2015). They found a handful of genes associated with early ripening which were significantly up-regulated including the 1-aminocyclopropane-1- carboxylate synthase (ACC synthase) encoding gene in the bud sports in comparison to their parent. 1.5.4. Genomic Understanding of Harvest Date in Apple Many studies have the genetic mechanisms linked to color change in apple bud sports (Cho et al., 2020; Iglesias et al., 2012; Y. Liu et al., 2022; H. Sun et al., 2023; Tian et al., 2022), but investigations into mutations resulting in altered harvest date have not been thoroughly explored. Regarding harvest date, several QTLs with a critical role in apple fruit maturation have been identified. The proposed fruit maturation date QTLs are numerous and complex, spread across 13 14 of the 17 chromosomes. The specific QTLs are found on linkage groups (LG) 3 (Migicovsky et al., 2016), 9 (Morimoto et al., 2013), 10, 15, and 16 (Kunihisa et al., 2014). Another finding suggests that there are 16 regions of the apple genome on 9 LGs linked to maturation date (Chagné et al., 2014). This leads to the idea that fruit developmental and/or maturation rate and the timing of apple harvest has very complex genetic underpinnings. For example, Morimoto et al. (2013) found correlations between fruit flesh anthocyanin content and harvest date. Also, fruit firmness and harvest date are thought to be tightly linked (Ban et al., 2022; Migicovsky et al., 2016). Growers find a tradeoff between early harvesting cultivars going to market earlier, but these cultivars do not store as long as later harvesting cultivars. 1.5.5. Gene Mutation Resulting in Altered Harvest Date The complexity of harvest date is discussed in the recent study that identified a candidate gene/genetic event for the mutation responsible for the late maturing ‘Autumn Gala’ phenotype (Ban et al., 2022). They utilized genomic and RNA-seq analysis throughout the growing season to analyze genetic alterations and differing gene expression. A genetic event called a retrotransposon insertion was discovered in this study. A retrotransposon insertion is when a segment of DNA is transcribed into RNA, reverse-transcribed back into DNA, and then inserted elsewhere in the genome. Results from Ban et al. (2022) suggest that a large (2.8 Mb) genomic deletion on chromosome 6 was caused by a 10.7 kb retrotransposon insertion. Because of the deletion, the remaining intact chromosome 6 had only one copy of each gene it contained, so these genes were the sole source of proteins from this chromosomal segment. In the late maturing phenotype ‘Autumn Gala’, a 10.5-fold suppression of ‘MdACT7’ resulted from a 2.5- kb insertion of a transposable element into that gene. Interestingly, both the parent and the mutant ‘Gala’ lines possessed the disrupted Mdact7 and it was suggested loss of the 2.8-Mb 14 segment with its functional version of MdACT7 resulted in the phenotype. The ‘MdACT7’ gene is orthologous to the ‘ACT7’ gene in Arabidopsis, an actin-related protein that plays a role in cell division, seed germination, root hair growth, and overall cell growth. Mutants of this gene display stunted growth and development phenotypes. However, apart from the rate of fruit development, the ‘Autumn Gala’ does not appear to differ phenotypically from the progenitor line. Since there is no other known phenotypic difference between ‘Autumn Gala’ and ‘Kidd’s D-8’, their finding is noteworthy. Ban et al. (2022) may be one of, if not the first, to explore the unique change in harvest date. Finding genomic variants is a vital aspect of discovering why these valuable mutations occur and evaluating the biology and physiology of the growth of both the apple tree and fruit bolsters genetic findings. 15 CHAPTER 2. Apple bud sport comparisons to progenitor/standard harvest cultivar of fruit growth suggest physiological predetermination of maturation date early in fruit development 2.1. INTRODUCTION 2.1.1. Background Apple bud sports contribute tremendous value to apple production. Bud sports originate from meristematic tissue in which a stable mutation has occurred. As the new genetic tissue arising from the mutated meristem continues to undergo cell division, differentiation, and expansion, new limb, leaf, and fruit phenotypes may emerge (Foster & Aranzana, 2018). Variable phenotypes in the fruit are most discoverable because of the distinct, visible difference in color, shape, or harvest date, as elaborated in the comprehensive review by Shamel and Pomeroy (Shamel and Pomeroy, 1936). Retold in LeAnn Zotta’s 200 Years and Growing: The Story of Stark Bro’s Nurseries & Orchards, the Stark family from Louisiana, Missouri first realized the value of bud sports when they discovered the bud sport they named ‘Starking® Delicious’. Their nursery began in the year 1816 and has continued producing nursery fruit trees ever since. ‘Starking® Delicious’ was originally discovered by New Jersey grower Lewis Mood, who noticed a limb producing red apples on the ‘Delicious’ (later marketed and known as the variety ‘Red Delicious’) tree while the rest were still green. Paul Stark Sr. rushed to pay $6,000 for this limb and now that limb has been grafted into millions of trees all over the world (Zotta, 2015). Ever since this event, apple bud sport numbers grew in value and popularity as a resourceful method of introducing and producing new apple cultivars. For example, within six years of the enacted Plant Patent Act (1930), over 1600 fruit tree bud sport patents were issued Shamel and Pomeroy, 1936). This means that there was at least one fruit tree patent issued every working day for six years. 16 Among the various classifications of bud sports in apple, ‘maturity sports’, sports that harvest at a different date than their progenitor, are of significant value. Early maturing sports bring more value than late sports, due primarily to their fruit’s early presence in the market, but also because of a shorter growing period and the potential for avoiding early autumnal frosts. The apple harvest window is an extremely important aspect of apple production. Each variety has a general harvest date window, varying year to year by plus or minus a couple of days, depending on environmental conditions (Shane et al., 2023). Growers spread their harvest window amongst different varieties to spread labor needs and diversify their crops. Growers also apply plant growth regulators days/weeks before harvest to slow ripening to delay harvest or retain fruit on the tree to prevent fruit drop (Bramlage et al., 1980). This practice is broadly utilized to improve labor efficiencies during harvest so field laborers can pick apples block by block in the orchard. 2.1.2. Fruit Development Fruit growth is a complex process that involves many metabolites and processes. In evaluating fruit development over the season, fruit size may be readily measured. Morphology during the development of apple fruits has been well characterized for decades (Tukey & Young, 1942) (Fig. 3). In apple fruits, cell division is the primary force behind fruit size for 7-10 days after bloom. After this first phase of fruit growth, cell expansion occurs simultaneously with cell division until about 4-6 weeks after bloom when cell division ceases (Lakso & Goffinet, 2017). From this point on, cell expansion and growth in intercellular space are the main contributors to fruit growth. In Michigan, the cool season promotes a sigmoidal curve of fruit growth (Tukey & Young, 1942) (Fig. 2), as opposed to the expo-linear pattern of fruit growth (A. Lakso & Goffinet, 2017). 17 Figure 2. Expolinear and sigmoidal curves showing volume increase of 3 different apple cultivars during the growing season. ‘Early Harvest’ exhibits an expolinear behavior while ‘Mcintosh’ and ‘Rome’ have a sigmoidal growth behavior. 18 Figure 3. Characterization of morphology in cultivar ‘Twenty Ounce’ across entire fruit development from 4 weeks before bloom until ripe. Anatomically distinct tissues named. 19 2.1.3. Photosynthesis in Apple There has been much research done on photosynthesis in apple leaves, but primarily in environmental (soil moisture, vapor pressure deficit, light, temperature) response and physiological (source and sink, shading, spur leaf vs. bourse leaf assimilation, carbon allocation) studies (Bhusal et al., 2019; Chun et al., 2002; DeJong, 2022; Fallahi et al., 2001; Grappadelli et al., 1994; Lakso, 1983; Schneider & Childers, 1941; X. Sun et al., 2018; Tartachnyk & Blanke, 2004; Wang et al., 2018). There is relatively little to no study, to the author’s knowledge, on differences in the carbon assimilation characteristics of bud sports when crop load has been managed carefully. Since the dawn of portable gas exchange systems, measurements in the field have not only become possible, but also more accurate and reflective of actual commercial growing conditions. In this experiment, we evaluated carbon assimilation on a diurnal basis, meaning that we took leaf measurements throughout the day, starting pre-dawn and ending after sunset. Leaf measurements across an entire day give a more accurate representation of total daily carbon assimilation as opposed to just measuring assimilation at just one time point during the day. With an analysis of daily carbon assimilation, we aimed to uncover any possible correlation between rate of gas exchange in leaf tissue and rate of fruit growth, development, and maturity. If higher carbon assimilation coincided with higher fruit growth rate, for instance, the more rapid fruit development may be enabled by a higher photosynthetic rate or, conversely, the photosynthetic rate might be enhanced by the demands of a higher growth rate. If developmental rate is driven by a higher carbon assimilation in earlier harvesting cultivars, one could look for differences in expression of genes heavily involved in carbon assimilation as a root cause of accelerated development. Alternatively, if no relationship exists between assimilation rates and 20 maturation date, then perhaps one might logically exclude genes related to photosynthesis as being causal for a shift in maturation date. 2.1.4. Experimental Design and Rationale For this study, we evaluated bloom date to uncover any differences in full bloom date. If we found differences in full bloom date, the direction of our search would go toward an analysis of flower development or carbohydrates in the beginning of the season. We also evaluated fruit development over the entire growing season until full fruit maturity and compared early and late sports using measures of fruit growth, rate of growth, acceleration of growth, and leaf photosynthesis. For our photosynthesis analysis, we measured carbon assimilation, but other parameters such as stomatal conductance and photosystem II efficiency were also measured. Lastly, we characterized the ripening behavior of each cultivar to determine if there is a pause or slowing in development before ripening of the later cultivars. Characterizing bloom, fruit development, carbon assimilation rates, and ripening among the three different comparisons was intended to narrow the window for our search for candidate genes and genetic events causing early or late maturation in bud sport mutations. 2.2. MATERIALS AND METHODS 2.2.1. Plant Material The apple trees analyzed in this experiment were located in a research plot at the MSU Clarksville Research Center (42°52’27.3”N 85°16’15.0”W). For each cultivar, three randomized blocks of 8 trees were planted, with tree spacings of 0.9 meters (3-feet) apart within rows, and 3.7 meters (12-feet) between rows with adequate water, nutrition, and pruning provided by the research station. The ‘Fuji’ and ‘Cripps Pink’ trees given by Philip Schwallier were planted in 2019 on ‘Bud-9’ rootstock. The ‘Kidd’s D-8’ and ‘Autumn Gala’ ‘scions were collected from 21 Catoctin Mountain Orchard (Thurmont, MD) by Dr. Christopher Gottschalk from the original tree from which the mutation in ‘Autumn Gala’ originated. The ‘Gala’ trees were grafted into Bud-9 rootstock in 2021. Five trees of each cultivar similar in architecture and tree trunk cross- sectional area were selected for fruit size, leaf photosynthesis, and maturity measurements. Additional trees were selected for DNA leaf tissue and RNA fruit tissue collection. 2.2.2. Bloom Bloom date was characterized by examining 10 trees randomly throughout the block per cultivar (five of which used for both fruit growth and photosynthesis analyses). ‘Full bloom’ was determined by visually estimating when 80% of all flowers on the trees were open. 2.2.3. Crop Load Management Crop load is an essential variable to consider in tree fruit because of its effects on fruit size, tree trunk growth, shoot number, node emergence, and leaf photosynthesis (DeJong, 2022; Palmer et al., 1997). The five trees of each cultivar that were selected for growth measurements were managed to achieve the same final fruit set for the growing season. Tree trunk cross-sectional area (TCA) in cm2 was measured 30 cm above the graft union of each tree. A multiplier for each variety was used to calculate the total crop load that the tree could handle to obtain maximum growth potential. The calculation is TCA multiplied by 9 for ‘Gala’, and a factor of 7 for ‘Fuji’ and ‘Cripps Pink’. ‘Fuji’ and ‘Pink Lady’ were given lower multipliers due to their biennial bearing nature. After the calculation in ‘Gala’, the target crop load was halved due to the trees’ young age. Immediately after petal fall, fruitlets were initially thinned by hand. Most fruitlets of all varieties were 5-9 mm in diameter at the time of thinning. Fruit was first thinned down to a king fruitlet every other cluster. After this, each tree was further thinned based on ‘Equilifruit disc’ (National Institute for Agricultural Research, 147 Rue de I’Université 75338 Paris Cedex) 22 recommendations. The ‘Equilifruit disc’ has notches to estimate the number of fruit that should remain on a branch based on circumference (in mm) of each limb. Each tree was thinned to a fruit number slightly above the crop load target to ensure that enough fruit would remain to achieve target crop load after ‘June drop’, a natural phenomenon when trees shed fruit (Larson & Kon, 2021). Fruit number was re-counted after ‘June drop’ and then thinned to the exact crop load target (7 fruit/tree trunk cross-sectional area (TCA) for ‘Gala’ and ‘Pink Lady’ and 6 fruit/TCA for Fuji) for the remainder of the growing season. On each tree, five spurs (one king fruit/spur), were selected for fruit diameter measurements during the summer of the year 2023. Clusters that were selected were located on ‘dards’ which are one-year old limbs that are approximately 1-4 inches in length, where the abundance of dards varies based upon vigor, pruning, and cultivar (Boyes, 1923). The selected dards were located on the outer canopy of the tree, having more exposure to sunlight and were the same limbs from which we gathered carbon assimilation data. Measurements were performed twice every week until the rate of growth slowed in the later portion of the season, at which time measurements were taken at least once per week. For the first five weeks post bloom, the apple fruitlet diameter was measured using a standard digital caliper. After about five weeks, measurements were conducted via a produce measuring gauge (Cranston Machinery Co., Oak Grove, Oregon). This gauge has an extendable metal strap that wraps around the fruit equator, giving a more accurate representation of the fruit diameter. The five fruit per tree that were measured for the entire season were harvested and taken to the lab for harvest maturity analysis when comparable fruit suggested fruit ripening had begun (see Fruit Maturity Analysis below). Growth curves as a function of growing degree hours (GDH) were generated using specialized curve-fitting software 23 (TableCurve 2.0, Jandel Scientific, San Rafael, CA). We used the Weibull (#8088) equation (Eq. 1) 𝑦 = 𝑎 + 𝑏(1 − exp (−((𝑥 + 𝑑 ∗ ln (2)# )! ! " )) (1) where x=GDH value (Table 1). Variables a, b, c, d, and e all change from one curve fit for a specific fruit to another, but every individual fruit’s data was fitted to the generated curve of the same equation with slightly different values for each variable according to small differences in growth behavior of each fruit. Growth rate (1st derivative of growth curve) and acceleration of growth (2nd derivative of growth curve) were manually calculated for each fruit throughout the growing season to identify peak acceleration, peak growth rate, and trough deceleration of fruit growth. These developmental milestones were used as targets for fruit tissue for RNA extraction and analysis (described in Chapter 3). For each date of diameter measurement, a ‘Student’s T- test’ was performed for volume, growth rate, and acceleration of fruit growth. Five trees per cultivar were considered replicates and the 5 fruits measured per tree were considered subsamples. Data for subsamples were averaged and the averages subjected to statistical analysis to compare mutant and standard cultivars. 24 Table 1: Fitted curve equation for fruit growth of each cultivar showing variables of Weibull ! 𝑦 = 𝑎 + 𝑏(1 − exp (−((𝑥 + 𝑑 ∗ ln (2)# ) )) where x=GDH value. These curves are equation depicted in growth rate and acceleration of growth figures 3-11. ! " Variables Kidd’s D-8 a b c d e R2 -14.1656 243.15592 28605.07 167182.7 15.4026 0.9995 Autumn Gala -14.1674 234.23149 29912.333 366756.62 33.08485 0.9996 September Wonder Fuji -14.5170 360.12959 30418.000 73339.687 6.063213 0.9997 Aztec Fuji Maslin -5.2252 391.10507 39198.310 53966.135 2.942034 0.9996 -3.4059 323.74065 41548.613 52884.155 2.392670 0.9998 Cripps Pink -25.5178 257.20485 37093.037 222948.29 11.69927 0.9997 25 2.2.5. Photosynthesis Measurements Diurnal photosynthesis measurements (i.e., measurements made periodically throughout the day from dawn until dusk) were taken using a portable photosynthesis instrument (LI 6800, LI-COR Biosciences, Lincoln, Nebraska). Carbon assimilation, stomatal conductance, photosystem II operating efficiency, and light intensity were measured. Dates were selected beginning at early fruit growth stage during late cell division, mid-season cell expansion, peak fruit growth rate, and ~2-3 days before and after harvest. The type of leaf that was measured on each tree was the first or second bourse shoot leaf that emerged from a bourse shoot subtending a king fruit. The bourse shoot leaf was selected because bourse shoots proportionately contribute more carbon resources to the fruit than the spur leaves (Wünsche & Lakso, 2000). For the first three time points during the growing season, diurnal measurements for each cultivar were taken on the same day. Diurnal measurements were also taken for each individual cultivar a few days prior and after the respective cultivar’s harvest date; these dates differed for each standard and mutant strain. On each date, bourse leaf photosynthesis parameters were measured for all trees within a one-hour period for each of six timepoints. The diurnal measurements were taken pre-dawn, 9 a.m., 12 p.m., 3 p.m., 6 p.m., and post-dusk. The initial and end-of-day measurement times changed as daylength shifted across the season. To select leaves in regions experiencing maximal sun exposure, the leaves were measured on the east side of the tree for the first 3 time points of the day and on the west side for the last 3 time points. We used publicly shared software (R, R Core Team, 2017) for analyses and plots. Total net carbon assimilation of the day was calculated as the area under the curve less the respiratory carbon emitted at night. Net carbon assimilation was fitted to a linear model with cultivars, development, and time of the day as fixed effects 26 using the ‘lm()’ function. Treatment differences were estimated using ‘emmeans()’ in the ‘emmeans’ package (Lenth et al., 2024). 2.2.6. Fruit Maturity Analysis During the latter part of the season, apples from each cultivar were evaluated every 3-4 days beginning approximately two weeks before anticipated commercial harvest, at commercial harvest (when the starch index reached 4 and when growers typically harvest for maximum storage and shelf-life quality) and continuing for approximately two weeks after the commercial harvest date. For the before and after harvest timepoints, fruit for maturity analysis came from non-crop load-managed trees. At harvest, the fruit used for tracking growth rate measurements were used for maturity analyses. For all maturity timepoints (excluding the harvest time point when the 25 tracked fruit were analyzed), 10 apples were placed into trays, imaged, and analyzed in the following parameters: weight, DA meter (chlorophyll absorbance), percent redness (subjective), background color (scale 1-5 shades of green to yellow-subjective), internal ethylene, firmness, starch index (scale 1-8-subjective), and titratable acidity. 2.2.7. Fruit Maturity Analysis Methodology Each apple was placed onto a scale and weight in grams recorded. Chlorophyll (DA) readings were taken from opposing sides of the fruit (red and green) of the apple. Each apple was analyzed by holding the fruit flush to a specialized hand-held reflectance spectrometer (DA Meter, Sintéleia, Turoni, University of Bologna-Department of Agricultural Science, AGRIMAT S.R.L, Bologna, Italy). The DA Meter measures the amount of chlorophyll in the apple by measuring its absorbance and outputting that absorbance as an index. Percent redness as a subjective measure was assessed by holding one’s index finger over and around the stem bowl, and the thumb under and around the calyx end of the apple. Both sides of the apple were 27 evaluated. Often the bi-colored apples such as ‘Gala’ and ‘Cripps Pink’ were analyzed by first evaluating the reddest side of the apple and then switching to the opposite side of the fruit. The percentage of redness of each side of the apple was estimated in increments of 5% and the average percentage for the two sides recorded. A background color index with 5 shades of green (5=dark green and 1=yellow) according to the ‘Macintosh’ cultivar was used (L.R. Simons, Cornell University Extension-New York State College of Agriculture, Ithaca, New York). The shade of color most accurately representing the background color of each fruit was recorded. A gas chromatograph was used to measure the internal ethylene concentration of each individual apple. A standard of 1 L L-1 ethylene was used (Matheson Tri-Gas, Inc.) and the sample volume was 1 mL. Fruit internal ethylene was measured by withdrawing a 1-mL gas sample from the core of the apple fruit. To do this, a needle, with a clean-out wire inserted to prevent the needle from becoming clogged by apple cortex/juices, was pushed into the calyx end of the fruit, entering the seed cavity of the fruit. The clean-out wire was removed from the needle and a 1-mL plastic syringe was mounted on the needle and 1 mL of the gas sample was extracted. The gas sample was then inserted into a gas chromatograph (Carle AGC series 400, Carle Instruments Company, Fullerton, CA) and the output recorded on a chart recorder (Linear 1200, Barnstead Thermolyne Co. Ramsey, MN, USA). A penetrometer (QA Supplies, FT327) attached to a portable manual drill press was used. A 2-cm dia. disc of peel was removed on opposing sides of the apple equator to expose the cortex. The apple was then held on the press platform and penetrated by the penetrometer to a depth of 1 cm. The penetrometer force was read in pounds and converted to Newtons. For measuring starch index, apples were sliced in half at the equator and one half was dipped into an iodine solution (8.8 grams of potassium iodide dissolved in 30 mL warm water followed by the dissolution of 2.2 grams of iodine into the solution, which was 28 then brought to 1 L volume). Each stained half was placed into a 20-cell tray and analyzed by comparison to a standardized starch index on a 1-8 scale (Cornell University Starch Index, Blanpied & Silsby, 1992). A hand-held refractometer (ATAGO CO., LTD, Tokyo, Japan) was used to analyze the sugar concentration within the fruit. A drop of juice from the apple was put onto the glass of the refractometer and the plastic cover sealed the juice onto the glass. A digital titratable acidity meter (ATAGO CO., LTD, Tokyo, Japan) was used to calculate the titratable acidity of each fruit. 1-mL of apple juice was placed into a 100-mL beaker. It was then diluted with 49 mL of deionized water and thoroughly mixed. A few drops of the resulting solution were then pipetted onto the digital refractometer/acidity sensor surface and the titratable acidity value was calculated. 2.3. RESULTS 2.3.1. Bloom There were no differences in bloom between the early and later harvesting cultivars in all three comparisons of the progenitor and sport (Fig. 4, Table 2). 29 Figure 4: Bloom depicted on May 10, 2023. No visible differences in bloom phenology were observed between maturity sports and standard cultivars. 2.3.2. Fruit Growth, Rate, and Acceleration of Fruit Growth Throughout the 2023 growing season, fruit diameter was measured at least once per week for all cultivars starting at 14 days after full bloom (DAFB). The R2 value for curves describing fruit volume as a function of GDH had values above R2=0.99, with most curves with value R2 =0.999 or above (Appendix). The fitted variables for the curves using the averages for each cultivar are presented (Table 1). All cultivars had comparable initial volumes (~0.3-0.5 cm3). Both the ‘Gala’ and ‘Fuji’ mutant/standard comparisons ended in similar final fruit size, but in the ‘Pink Lady’ comparison, ‘Cripps Pink’ had less volume at final fruit size than 'Maslin Cripps Pink'. The once-per-week measurement of fruit diameter enabled the collection of highly precise data for curve fitting. This precision enabled a more reliable determination of the first date of divergence in fruit growth between the progenitor and the sport than fruit volume measurements. 30 In ‘Gala’, the later harvesting bud sport ‘Autumn Gala’ had a smaller volume than its progenitor ‘Kidd’s D-8’ by 17 July, 67 DAFB (Fig. 5). Analysis of fruit growth rate revealed that ‘Autumn Gala’ grew at a significantly lower rate than ‘Kidd’s D-8’ by 15 June, 35 DAFB, which was more than a month before differences were evident in fruit volume (Fig. 6). The fruit growth acceleration of ‘Autumn Gala’ was significantly lower than its progenitor by 1 June, 21 DAFB, which was two weeks prior to the first day of growth rate separation between the cultivars (Fig. 7). Figure 5: Calculated fruit volume in ‘Gala’ cultivars over the entire season based on measures of fruit diameter or circumference and assuming spherical fruit shape. The fruit volume of late bud sport ‘Autumn Gala’ was less than its progenitor after 17 July, 67 DAFB and 23,509 GDH (arrow; T-test-p=0.039). Each data point represents the average fruit size of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. ‘Kidd’s D-8’ R2=0.9994 and ‘Autumn Gala’ R2=0.9996. 31 Figure 6: Fruit growth rate in ‘Gala’ over the growing season. The growth rate in late bud sport ‘Autumn Gala’ differed initially from its progenitor on 15 June, 35 DAFB and 10,329 GDH (arrow; T-test-p=0.050). Each data point represents the average growth rate of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. 32 Figure 7: The 2nd derivatives of the data shown in Figure 5. Acceleration of fruit growth was significantly lower in ‘Autumn Gala’ beginning on 1 June, 21 DAFB and 5,993 GDH (arrow; T- test-p=0.044). Each data point represents the average growth acceleration for 25 fruits from a total of 5 trees (5 fruits per tree) on a date of in-field measurement. Vertical bars represent standard error of the averages of each tree replicate. In ‘Fuji’, the early sport ‘September Wonder Fuji’ had a greater fruit volume than the standard cultivar ‘Aztec Fuji’ by 8 July, 59 DAFB (Fig. 8). ‘September Wonder Fuji’ had a higher growth rate than ‘Aztec Fuji’ by 25 May, 15 DAFB (Fig. 9). ‘September Wonder Fuji’ had a significantly higher fruit growth acceleration by 25 May, 15 DAFB, compared to ‘Aztec Fuji’ (Fig. 10). ‘September Wonder Fuji’ had a higher rate of maximum growth acceleration at ~18,000 GDH, 3 July, than ‘Aztec Fuji’. ‘September Wonder Fuji’ had a faster and earlier deceleration than ‘Aztec Fuji’, with the maximum rate of deceleration occurring at ~45,000 GDH, 8 September. The maximum rate of deceleration in ‘Aztec Fuji’ was not as high as ‘September Wonder Fuji’ and occurred about 10,000 GDH later, on 24 October. 33 Figure 8: Calculated fruit volume in ‘Fuji’ cultivars over the entire season based on measures of fruit diameter or circumference and assuming spherical fruit shape. The fruit volume in early bud sport ‘September Wonder Fuji’ was greater than its standard comparator after 8 July, 59 DAFB and 20,166 GDH (arrow; T-test-p=0.036). Each data point represents the average fruit volume of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. ‘September Wonder Fuji’ R2=0.9997 and ‘Aztec Fuji’ R2=0.9996. 34 Figure 9: Fruit growth rate in ‘Fuji’ cultivars over the growing season. The growth rate in early bud sport ‘September Wonder Fuji’ was higher than its progenitor by 23 May, 13 DAFB and 3,917 GDH (arrow; T-test-p=0.0094). Each data point represents the average growth rate of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. 35 Figure 10: The 2nd derivatives of the data shown in ‘Figure 8’. Fruit volume acceleration was initially higher in ‘September Wonder Fuji’ by 23 May, 13 DAFB and 3,917 GDH (arrow; T-test p=0.041). Each data point represents the average fruit growth acceleration of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. 36 In ‘Cripps Pink’, the early harvesting sport ‘Maslin Cripps Pink’ had a greater fruit volume than the standard harvesting cultivar ‘Cripps Pink’ by 29 July, 80 DAFB (Fig. 11). The fruit growth rate of ‘Maslin Cripps Pink’ was significantly higher than ‘Cripps Pink’ by 5 June, 26 DAFB (Fig. 12). Interestingly, the initial growth rate of ‘Cripps Pink’ was higher than ‘Maslin Cripps Pink’, but by 5 June, 26 DAFB, ‘Maslin Cripps Pink’ had already overtaken ‘Cripps Pink’ in growth rate. Acceleration of fruit growth was significantly higher in ‘Maslin Cripps Pink’ by 25 May, 15 DAFB (Fig. 13). Figure 11: Calculated fruit volume in ‘Pink Lady’ cultivars over the entire season based on measures of fruit diameter or circumference and assuming spherical fruit shape. Significantly higher volume in early bud sport ‘Maslin’ was greater than its progenitor after 29 July, 79 DAFB and 28,897 GDH (arrow; T-test-p=0.031). Each data point represents the average fruit volume of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. ‘Maslin’ R2=0.9997 and ‘Cripps Pink R2=0.9997. 37 Figure 12: Fruit growth rate in ‘Pink Lady’ cultivars over the growing season. Significantly higher growth rate in early bud sport ‘Maslin’ on 5 June, 26 DAFB and 7,848 GDH (arrow; T- test-p=0.0303). Each data point represents the average fruit growth rate of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. 38 Figure 13: The 2nd derivatives of the data shown in ‘Figure 11’. Acceleration of fruit growth significantly higher in ‘Maslin’ on 23 May, 13 DAFB and 3,917 GDH (arrow; T-test-p=0.0003). Each data point represents the average fruit growth acceleration of 25 fruits from a total of 5 trees (5 fruits per tree). Vertical bars represent standard error of the averages of each tree replicate. Generally, fruit volume differed between the progenitor and the sport approximately halfway through the growing season. However, when the rate of growth was calculated based on the fitted fruit volume curves, the progenitor and the sport phenotype differed much earlier in development in the exponential phase of the growth curve in each comparison. Differences in fruit growth acceleration was also detected early in each progenitor/sport comparison in the exponential, early phase of volumetric fruit growth. In conclusion, differences in the rate of development can be detected between early and late maturing cultivars at the earliest stages of development and early maturity sports develop at a faster rate than the later maturity sports. 39 2.3.3. Photosynthesis Net carbon assimilation for all six cultivars on five days throughout the season was determined by calculating the integral of each diurnal assimilation curve for each leaf measured per cultivar and tree. Generally, the rate of assimilation did not change dramatically between mid-June and mid-August, averaging approximately 0.75 mol m-2 d-1 for the 'Fuji' strains and 0.6 for the 'Gala' and 'Pink Lady' strains. The ‘Fuji’ and ‘Pink Lady’ cultivars had similar carbon assimilation rates during the growing season until harvest. ‘September Wonder Fuji’, the early sport, matured approximately 27 d before ‘Aztec Fuji’ yet had a lower carbon assimilation rate than ‘Aztec Fuji’ at the developmental preharvest timepoint, 3 days before harvest (Fig. 14). The assimilation rates at harvest declined relative to the mid-season averages due likely to temperature and light decline later in the season (Fig. 15). Significant decreases of CO2 assimilation after harvest were exhibited by each cultivar except early bud sport ‘September Wonder Fuji’ and standard harvesting ‘Cripps Pink’. 40 Figure 14: Net carbon assimilation rates for bourse shoot leaves adjacent to apple fruit on 16 June, 27 July, and 18 August and 2 to 5 days pre- and 2 to 5 days post-harvest for the six cultivars evaluated. Total carbon assimilation calculated for each day as the average integral under the curve of each tree containing two subsamples. Pre- and post-harvest measurements were based on the fruit achieving commercial maturity. Each data point represents total carbon assimilated during that day. 2 leaves were measured per tree, 5 trees were measured, and measurements were taken 6 times sequentially throughout the day from dawn until dusk. Vertical bars represent standard deviation. Asterisks indicate significant difference (p≤0.05). 41 Figure 15: Growing degree hour (GDH) accumulation and solar flux per day according to date from 16 June 2023 through 9 November 2023, spanning all photosynthetic rate assays. Both GDH accumulation and solar flux decrease from 16 June to 9 November. Solar flux was calculated from Clarksville Research Center from ‘Langley’ units to kJ/m-2s-1 for each photosynthetic rate assay date. 42 2.3.3. Fruit Maturity There were stark differences in the timing of changes in maturation and ripening indices between different cultivars and between standard/progenitor lines and their respective sport lines. The starch index data inform us that not only is harvest date different, but starch conversion also begins at different dates (Fig. 16). Thus, the delay we see in harvest date is not due to a shortened or prolonged maturation period, but rather due to an advancement or delay in the development of the fruit leading to an advancement or delay in reaching physiological maturity. Data for other indices of maturity and ripening [e.g., weight, internal ethylene content, percent redness, firmness, background color, chlorophyll absorbance, percent sugar (°Brix), and percent malic acid (titratable acid)] (Figs. 17-19) are consistent with the starch conversion data, and further establish the clear difference in time of maturation between the early and late maturing cultivars. Harvest dates were determined by tracking ripening, primarily starch index, approximately 2 weeks before predicted harvest of each cultivar until a few days after each cultivar had an average starch index of 4 (Table 2). Table 2: Bloom, harvest dates, days from full bloom until harvest (DTH) and growing degree hours (GDH) at harvest. Harvest dates were determined as the date associated with attaining a starch index of 4 according to the Cornell University Starch Index (Blanpied & Silsby, 1992). Apple Variety Timing Cultivar Full Bloom Date Harvest Date DTH GDH 'Gala' 'Fuji' 'Pink Lady' Standard 'Kidd’s D- 8' Late 'Autumn Gala' Early 'September Wonder' Standard 'Aztec Fuji' Early 'Maslin' Standard 'Cripps Pink' May 11 May 11 May 10 May 10 May 10 May 10 Sep. 6 118 44,002 Sep. 30 Sep. 6 October 4 October 27 Nov. 5 142 51,221 119 44,293 147 53,120 170 56,301 179 56,753 43 Figure 16: Starch index analysis of all six cultivars. Commercial harvest date for each cultivar was determined by a starch index of 4 according to the Cornell University Starch Index (Blanpied & Silsby, 1992). 44 Figure 17: Maturity indices for ‘Kidd’s D-8’ and ‘Autumn Gala’ fruit from two weeks prior to two weeks after harvest date. Harvest date was 6 September (118 DAFB) for ‘Kidd’s D-8’ and 30 September (142 DAFB) for ‘Autumn Gala’ based on starch index. Each data point represents an average of 10 fruits randomly selected from 5 similar trees. Vertical lines represent standard deviation. 45 Figure 18: Maturity indices for ‘September Wonder Fuji’ and ‘Aztec Fuji’ fruit from two weeks prior to two weeks after harvest date. Harvest date was 6 September (119 DAFB) for ‘September Wonder Fuji’ and 4 October (147 DAFB) for ‘Aztec Fuji’. Each data point represents an average of 10 fruits randomly selected from 5 similar trees. Vertical lines represent standard deviation. 46 Figure 19: Maturity indices for ‘Maslin’ and ‘Cripps Pink’ fruit from two weeks prior to two weeks after harvest date. Harvest date was 27 October (170 DAFB) for ‘Maslin’ and 5 November (179 DAFB) for ‘Cripps Pink’. Each data point represents an average of 10 fruits randomly selected from 5 similar trees. Vertical lines represent standard deviation. 47 2.4. DISCUSSION 2.4.1. Value of Germplasm Background The main objective of this study was to identify and characterize differences in fruit development in apple bud sport cultivars that harvest at different dates than the standard or progenitor line. An important variable to consider is that two of the three comparisons were to the bud sport’s progenitors. In the case of ‘Gala’, the comparison is direct, in that the tissue grafted and used in this study was taken directly from the parent tree and its mutant limb that gave rise to the 'Autumn Gala' and differs in this regard from the study done by Ban et al. (2022). The reason that this is relevant is that fruit trees are known to accumulate somatic mutations at a relatively high rate (Sun et al., 2023). Thus, there is the risk that additional genetic alterations may have accumulated in lines that have undergone multiple cloning cycles. In ‘Pink Lady’, although our comparison is not direct, meaning we don’t have the original germplasm of the mutation, the standard harvesting ‘Cripps Pink’ is still the progenitor of bud sport ‘Maslin’ and thus a valuable comparison. These direct comparisons are valuable because it is possible that there was very likely only a single mutation responsible for the delay or advancement of maturity in the bud sports. This is not to say that only a single gene is responsible. In the case of deletion events, large portions of chromosomes may be lost or disrupted. Thus, characterizing fruit development differing by strictly one mutation or genetic event may be valuable in further understanding apple fruit development and maturity determination. 2.4.2. Earlier Harvesting Apple Cultivars Exhibit Compressed Developmental Periods The higher fruit growth rates of earlier harvesting lines relative to the later lines during the exponential phase of volumetric fruit growth suggest genetic differences in the early and late lines is the result if shifts in physiology apparent immediately after flowering and are therefore 48 likely not related to maturation rate, per se. This may result from faster cell division rates, altered hormonal regulation, shifts in the duration of the cell division phase, or a combination of these processes. The data suggest that maturation and ripening rates were similar between early and late cultivars for each of the three varietal lines. The difference between the early and late cultivars regarding maturity was the timing of the onset of maturation and ripening, not the rate of either process. If the early and late cultivars of each comparison had similar dates of full maturation with a difference in ripening rate, the difference in maturity would thus be primarily attributed to a difference in ripening rate and not a compressed development. Since we see a significantly higher rate of fruit growth in all the earlier harvesting cultivars and simultaneously observe earlier harvest dates with no difference in ripening rates, we suggest that the physiological events leading to earlier harvesting cultivars is not due to differences in the physiology associated with ripening, but likely have a compressed developmental period before full maturity. Thus, genetic changes that have occurred in the early and late sports should be a result of gene expression shifts detectable in the early stages of fruit development. So, while bloom date may not differ (Table 2), future genetic expression analyses should target early stages for signs of differentially expressed genes important in the developmental process causing early/late maturation. 2.4.3. Increase or Decrease in Net Carbon Assimilation Not Responsible for Faster Development in Earlier Harvesting Cultivars Since the rate of carbon assimilation by bourse leaves was not uniformly higher for the earlier cultivars, data suggest that earlier harvest is likely not driven by a higher carbon assimilation rate during the early and mid-growing season fruit growth. In the ‘Gala’ comparison for example, we would expect that the earlier harvesting cultivar ‘Kidd’s D-8’ would assimilate more carbon 49 earlier in the season. On the contrary, ‘Kidd’s D-8’ assimilates significantly less carbon. The preharvest carbon assimilation of the early cultivars ‘Kidd’s D-8’ and ‘September Wonder Fuji’ are both lower than the carbon assimilation of their later harvesting progenitor/standard harvest time cultivar. An observation worth mentioning, not relevant to bud sport origin or characterization but rather differences among apple varieties, is that ‘Fuji’ achieved a higher rate of carbon assimilation thank ‘Gala’ or ‘Cripps Pink’ during the growing season, especially on July 27. The final fruit size of both ‘Fuji’ cultivars was also found to be significantly higher than the ‘Gala’ or ‘Pink Lady’ cultivars. A more complete study providing a comprehensive analysis for the full canopy should provide needed details to understand how shifts in the crop's earliness or lateness alter carbon assimilation, flux, and emission. 2.4.4. Relevance of the Study There has never been comprehensive work done on apple growth rate analysis between apple sports within the same commercial variety, let alone for the purpose of uncovering mechanisms underlying mutated maturity phenotype. Many molecular studies in horticulture lack a robust physiological element in the study. In a recently published paper on a study similar to ours, Ban et al. (2022) found a candidate gene (MdACT7) they proposed may cause late maturation in ‘Autumn Gala’ compared to ‘Kidd’s D-8’. They measured apple fruits throughout the season but never mentioned crop load requirements of the trees or did any further analysis of fruit growth rate/acceleration of fruit growth. Dong et al. (2011) evaluated early ripening events in ‘Beni Shogun’ (‘Fuji’) and heavily characterized the earlier ethylene burst in ‘Beni Shogun’ and looked at other ripening characteristics such as volatile production, skin coloration, fruit softening, and starch hydrolysis. In this study, only ripening was extensively evaluated; no preharvest physiological assays were performed. Kim et al. (2023) also evaluated ripening 50 behavior in early cultivar ‘Beni Shogun’ compared to ‘Fuji’ (Kim et al., 2023). Again, season- long developmental analyses were not performed. Studies focusing on ripening behavior is not surprising, as the phenotype only portrays itself late in development. In analyzing our data, signs of phenotypic differences present themselves very early in apple fruit development. Genomic and transcriptomic analyses are even more powerful when coupled with comprehensive characterization of physiological analyses such as rate of fruit development, rate of carbon assimilation, and ripening behavior. These analyses provide a faster approach to transcriptomic studies due to the narrow window we identified in early development when these cultivars diverged in developmental rate rather than continue to evaluate ripening behavior and gene expression during ripening. Further, given the lack of a consistent relationship between photosynthetic carbon assimilation and earliness or lateness of the cultivar comparisons, the search for the genetic underpinnings controlling maturity timing appear to be less likely to involve genes directly involved in assimilation. We characterized each of the six cultivars, 3 variety pair-comparisons, from bloom until postharvest ripening, and this characterization may lead us and others to a much clearer understanding of the likely complex genetic control of harvest date in apple and maturity bud sport origin. 51 CHAPTER 3. Genetic underpinnings of early and late maturing apple bud sports 3.1. INTRODUCTION Apple bud sports are spontaneous mutations in meristematic tissue such that a stable somatic mutation develops through a whole limb, flower, and/or fruit (Foster & Aranzana, 2018). This phenomenon leads to a natural process that humans may take advantage of to produce higher quality fruit. Bud sports that are commercialized into widespread cultivation often contain all desirable attributes of the parent tree, with one, sometimes more, additional quality attributes that justify the high cost of orchard establishment for the genetically novel material. Common bud sport characteristics include change in fruit size, shape, color, spur-bearing behavior, and advanced or delayed harvest. Harvest date variation is an important concept to the apple industry. Spreading out the harvest window and diversifying apple varieties in the orchard contributes to a growers’ productivity. In the case of altered harvest date apple bud sports, hereafter referred to as ‘maturity sports’, apple fruits may reach their optimal harvest date at an earlier or later date than their progenitor. Growers often take advantage of early bud sports due to their earlier presence in the market. Later bud sports may be advantageous due to higher fruit firmness and thus storability since higher storability and harvest date are thought to be heavily correlated (Ban et al., 2022; Migicovsky et al., 2016). Later cultivars may also be used to prolong the growing season for higher apple consumption availability later in the season. Although these valuable maturity sports are frequently utilized in the apple industry, the physiological and genetic mechanism behind the phenotype of early or delayed maturation date in maturity sports is currently unknown. Importantly, a proposed genetic link to late fruit maturation in ‘Gala’ has been 52 proposed (Ban et al., 2022), but conclusive data on the mechanism of this exciting finding are still lacking. Molecular studies of bud sports have varied in their approach from observing thousands of genes and several gene groups separated by function, to finding one causal candidate gene responsible for the phenotype of early or late maturity. For example, in apple, Ban et al. (2022) compared the standard harvesting ‘Kidd’s D-8 Gala’ with its late maturing bud sport ‘Autumn Gala’ (Table 1). They suggested that a dysfunctional ACT7 gene (Mdact7) may be responsible for the alteration in phenotype of the later cultivar. They explored the function of Mdact7 in arabidopsis (Arabidopsis thaliana). In arabidopsis, an actin-7-like encoding gene ortholog is responsible for critical roles in plant development due to its effect on plant height and general stunting of the plant. Actin is a crucial protein in plant cells that exists in the form of actin filaments within the cytoskeleton of the cell and participates in multiple essential cellular processes such as cell expansion, cell division, and intracellular transport (Thomas et al., 2009). This ACT7 gene is the only actin ortholog in Arabidopsis that responds to exogenous hormones and external stimuli such as auxin and light regime and wounding, respectively (McDowell et al., 1996). Transgenic lines that contained the mutated allele Mdact7 found in ‘Autumn Gala’ exhibited stunted growth in arabidopsis. Maturity sports have also been found and studied in other horticultural crops such as citrus (Citrus sinensis), pear, (Pyrus bretschneideri Rehd.), and grape (Vitis vinifera L.). For example, in grape, Wei et al. (2020) found several differentially expressed candidate genes (Table 3) associated with hormone signaling and biosynthesis which may contribute to an early maturation phenotype. Liu et al. (2014) performed a proteomic analysis on the early maturing pear bud sport ‘Zaosu’ with and found proteins related to cell-wall modification, oxidative stress, 53 pentose phosphate metabolism, photosynthesis, glycolysis, and vital cellular processes were higher in abundance (Table 3). The search for mechanisms resulting in altered phenotypes in crops has narrowed from genetic regions to specific genes containing larger structural variants, small insertions, deletions, and single nucleotide polymorphisms (SNPs). Likewise, investigating the underlying mechanisms of bud sports differing in harvest date has, over time, become more refined, resulting in more specific findings (e.g., the identification of single candidate gene candidates that have been proposed or shown to contribute to maturity phenotypes as opposed to lists of genes and proteins that differ in expression). Regarding developmental time required for maturation, breeding efforts may soon be able to focus on generating cultivars for specific growing regions, which require shorter or longer growing seasons, using marker assisted selection. Our physiological work helps narrow marker assisted selection to early on in fruit development by reducing the type of genes to genes heavily involved in fruit development early in the season which will lead to a much clearer understanding of the complex genetic control of harvest date in apple and apple maturity bud sport origin. 54 Table 3: Summary of studies aimed at identifying causative mutations and molecular mechanisms for maturity sport phenotypes. Crop Apple (Malus domestica ) Apple (Malus domestica ) Apple (Malus domestica ) Sport Cultivar(s) ‘Autumn Gala’ Parent Publication ‘Kidd’s D-8’ Ban et al. 2022 ‘Hirosaki Fuji’ ‘Fuji’ Wang et al. 2009 ‘Beni Shogun’ ‘Yataka’ (early sport of ‘Fuji’) Kim and Ban et al. 2023 Pear (Pyrus bretschneideri Rehd.) *Unnamed early maturing bud sport ‘Zaosu’ Liu et al. 2014 Sport Phenotype 4-week maturation delay 40-day earlier maturation 3-week earlier maturation than Fuji Earlier maturation Candidate genes MdACT7 MdACS1 MdACO1 MdETR1 MdERS1 MdERS2 MdPG1 MdHSP17.5 MdACO1 MdARF1 MdIAA11 MdNAC3 MdNAC5 MdMADS7 MdMADS8 Several genes (proteomic analysis) Navel Orange (Citrus sinensis) Table Grape (Vitis vinifera L.) ‘Fengjiewan- cheng’ ‘Fengjie 72- 1’ Liu et al. 2007 ‘Tiangong Moyu’ ‘Summer Black’ Wei et al. 2020 CitSS1 CITAI CitCS CitAC ~45 DEGs 1 month maturation delay 10-day earlier maturation Functions of Candidate Genes Actin homolog involved in cytoskeleton, cell expansion and division, cellular transport Ethylene biosynthesis Ethylene receptor proteins Cell wall degradation Heat shock protein Ethylene biosynthesis Auxin regulation Fruit development and ripening transcription factors Cell-wall modification Oxidative stress Pentose phosphate metabolism Photosynthesis Glycolysis Sucrose synthase Acid invertase Mitochondrial citrate synthase Cytosolic aconitase Hormone signaling/biosynthesis Phenolic biosynthesis Flavonoid biosynthesis Anthocyanin biosynthesis Calcium response Plant-pathogen response Sugar accumulation Cell wall (GRIP) 55 3.2. MATERIALS AND METHODS 3.2.1. Plant Material The apple trees used for this research were located within a 1.2-acre trellised high-density planting at the Michigan State University Clarksville Research Center (Clarksville, MI 42°52’27.3”N 85°16’15.0”W). For each cultivar, three randomized blocks of 8 trees were planted, with tree spacings of 0.9 meters (3-feet) apart within rows, and 3.7 meters (12-feet) between rows. Budwood for the two ‘Fuji’ and two ‘Pink Lady’ cultivars was provided by Schwallier’s Country Basket (Sparta, Michigan). Budwood for the ‘Gala’ varieties was taken from the original ‘Autumn Gala’ sport branch and branches from progenitor ‘Kidd’s D-8’ tree at Catoctin Mtn Orchards (Thurmont, MD). ‘Fuji’ and ‘Cripps Pink’ budwood was bench grafted onto bareroot Bud-9 rootstock trees in 2019 and planted in the field on 13 May 2019. The ‘Gala’ trees were side-grafted onto two-year old Bud-9 rootstock on 27 April 2021. 3.2.2. Tissue Collections for DNA Sequencing and Future RNA Sequencing Leaf tissue of each cultivar was collected, flash frozen in liquid nitrogen in the field, and transported on dry ice before being stored in a -80 °C freezer. DNA was extracted from 3-5 leaves per tree for each cultivar. ‘Gala’ tissue collection was of newly emerging leaves from early summer growth on 26 May of 2022. ‘Fuji’ and ‘Pink Lady’ tissues were collected on 24 October of 2022 when there was no actively growing leaf tissue. Leaf tissue for the later cultivars was more aged and more difficult to extract DNA from. Four random trees of each cultivar similar in architecture and tree trunk cross-sectional area were selected randomly throughout the plot for transcriptomic analysis. For each apple variety, at least 10 timepoints were selected to capture important milestones in the developmental periods that included cell division, peak rate of change of fruit growth, peak fruit growth rate, lowest rate of change of fruit 56 growth, full fruit maturity, and ripening (Fig. 20). Fruit from all cultivars were collected on 23 May, 6, 13, 20 June, 3, 11, 18, 25 July, 1 August, 13 September. Additional collections were made for ‘Autumn Gala’ on 26 September, ‘Aztec Fuji’, ‘Maslin’, and ‘Cripps Pink’ on 13 and 31 October, and ‘Cripps Pink’ on 15 November. Of the 4 selected trees per genotype, 3 representative fruit were cut into pieces and put into tubes, then flash-frozen in liquid nitrogen. They were then transported on dry ice to the laboratory and stored in a -80 ℃ freezer until extraction. The first six fruit tissue collections were done slicing the whole fruit into similar sized cubes that would fit into a 50-mL tube. For the 5 later timepoints, tissue was separated between 4 different tissues: peel, cortex, core, and seed. This differentiated tissue was collected with the same methodology as the first 6 timepoints. Extraction of these first 6 timepoints was not complete at the time of assembly of this dissertation. 3.2.3. DNA Sequencing DNA was extracted from leaf tissue using the Qiagen DNEasy plant mini kit (QIAGEN, Germantown, Maryland). Each sample concentration was ≥ 5 ng µL-1. Illumina Next Generation library preparation and sequencing was performed by the RTSF Genomics Core at Michigan State University (project ID ENG13345). Libraries were prepared using the Roche Kapa HyperPrep DNA Library Kit with Unique Dual Index adapters following manufacturers’ recommendations. Completed libraries were quality checked and quantified using a combination of Qubit dsDNA HS and Agilent 4200 TapeStation HS DNA1000 assays. The libraries were pooled in equimolar quantities and this pool quantified using the Invitrogen Collibri Quantification qPCR kit. This pool was loaded onto one lane of an Illumina v1.5 SP flow cell using the Xp Workflow. Sequencing was performed in a '2x150 bp paired end format' using a NovaSeq 6000 v1.5 300 cycle reagent cartridge. Base calling was done by Illumina Real Time 57 Analysis (RTA) v3.4.4 and output of RTA was demultiplexed and converted to FastQ format with Illumina Bcl2fastq v2.20.0. 3.2.4. Variant Filtering Sequenced genomes of ‘Kidd’s D-8’, ‘Autumn Gala’, ‘September Wonder Fuji’, ‘Aztec Fuji’, ‘Maslin’, and ‘Cripps Pink’ were mapped to the ‘Gala’ haploid genome (Sun et al., 2020). The ‘Gala’ haploid genome was used primarily for ‘Gala’ due to higher accuracy of calls according to a closer related cultivar compared to other published genomes such as ‘Golden Delicious’ and ‘Honeycrisp’ (Daccord et al., 2017; Khan et al., 2022). After mapping, reads were analyzed for variants. For SNP calling, tools ‘deepvariant’ and ‘freebayes’ were used. For a variant to qualify, it needed to be called in both tools. In structural variant calling, program tools ‘TIDDIT’ and ‘manta’ were used. To narrow candidate variants with higher probability of having deleterious effects, structural variants needed to be at least 30 bp long, called by both tools, and coverage less than 3-4x higher than chromosome average coverage (depending on which tool). Although this filtering process eliminated smaller variants (<30 bp), it should be emphasized that small variants may also cause deleterious effects within the genome. The files for the large (structural variants) and SNP variants were established as follows (using the ‘Gala’ comparison as an example): (1) Same variants in both ‘Kidd’s D-8’ and ‘Autumn Gala’ but different genotype, meaning that where ‘Kidd’s D-8’ was heterozygous for a particular variant, ‘Autumn Gala’ was homozygous. The homozygous variant in ‘Autumn Gala’ refers to, as an example, a deletion occurring in a specific intragenic region on one chromosome of 'Kidd's D-8' and the same deletion occurring on both chromosomes of ‘Autumn Gala’. (2) A unique variant specific to a single cultivar (e.g. structural variant or SNP in ‘Autumn Gala’) that differed from both the 58 reference genome as well as the early or late cultivar to which the source is being compared. Variants in intragenic regions were selected for further evaluation. 3.3 RESULTS 3.3.1. Genetic Variant Identification Genomic DNA sequencing reads for all six cultivars were aligned to the annotated ‘Gala’ haploid genome (Sun et al., 2020). Coverage of haploid genomic data was relatively sufficient for confident analyses, considering 25x as optimal coverage (Table 4). At the time of the analysis annotated genomes for ‘Fuji’ or ‘Pink Lady’ cultivars were unavailable. Table 4: Coverage of each cultivar according to the ‘Gala’ haploid reference genome (X. Sun et al., 2020). Kidd’s D- 8 Autumn Gala September Wonder Fuji Aztec Fuji Maslin Cripps Pink Haploid 27x 25x 19x 19x 23x 22x Following alignments, structural variants (e.g., Insertions/Deletions) and SNPs between cultivars and the annotated genomes were identified. Variant calls were then compared between each maturity sport and the corresponding standard maturing cultivar of the same variety. The calls for variants in the tables were called heterozygous and homozygous for standard harvesting cultivars and bud sports, respectively. This means that when a variant was shared between both cultivars in the comparison, the bud sport lost the copy or function of the gene where the variant was called. In the tables containing unique variants, variants were filtered upon two important factors: (1) they were called within an intragenic region which would likely disrupt the coding regions (e.g. missense, frameshift, gain or loss of start or stop codon) and (2) called homozygous by both software programs that we used (‘manta’ and ‘tiddit’). This means that all unique 59 variants had no alternative allele to call upon for a specific gene where the variant was located. This later filtering method should reveal the most likely candidates for change in maturation. 3.3.2. DNA Variants in ‘Kidd’s D-8’ and ‘Autumn Gala’ from Haploid Genome Alignments Variant calling for the ‘Gala’ cultivars using the haploid mapping identified 15 SNPS and four SVs as potential causative mutations based on our selection criteria (Tables 5, 6, 7, 8, 9, 10). Surprisingly, 11 of the SNPs were in genes in a 2.57 Mb region on chromosome 6 between bases 21,828,353 and 24,402,010 and were all heterozygous in ‘Kidd’s D-8’ and homozygous in ‘Autumn Gala’, indicating a loss of heterozygosity in the late sport (Table 5). This region also contained two structural variants that were heterozygous in ‘Kidd’s D-8’ and homozygous in ‘Autumn Gala’ (located in Mdg_06g010800 starting at base 21,907,728 and Mdg_06g011660 starting at base 23,072,809) (Table 6). A deletion unique to ‘Autumn Gala’ was identified in this region as well (starting at base 22,686,399 in Mdg_06g011460). These variants span a region of chromosome 6 that maps to the location of the previously reported 2.8 MB deletion (between 24.913 to 27.743 Mb) and the gypsy retrotransposon 10.7 KB insertion associated with the ‘Autumn Gala’ phenotype (Ban et al., 2022). However, no variant was called for the ortholog of ACT7 in our variant analysis. Additionally, we found other orthologs of ACT7 within other regions of the ‘Gala’ genome (Chr. 14 and 15). 60 Table 5. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where, when called heterozygous for ‘Kidd’s D-8’, ‘Autumn Gala’ SNPs were called homozygous. Chromosome Gene Variant Type Kidd’s D-8 6 6 6 6 6 6 6 6 6 6 6 Mdg_06g010730 Mdg_06g010970 Stop codon gained Stop codon gained Mdg_06g011060 Mdg_06g011240 Mdg_06g011320 Mdg_06g011580 Mdg_06g011670 Mdg_06g011710 Mdg_06g011710 Mdg_06g012310 Mdg_06g012630 Splice acceptor variant and intron variant Splice donor variant and intron variant Frameshift variant and synonymous variant Stop lost and splice region variant Frameshift variant Frameshift variant Frameshift variant Splice donor variant and intron variant Frameshift variant Stop codon lost 13 Mdg_13g005100 Genotype Heterozygous Autumn Gala Genotype Homozygous Variant location Gene ortholog Gene Description 21828353 N/A Uncharacterized protein Heterozygous Homozygous 22031482 AT1G64550 ABC Heterozygous Homozygous 22148302 N/A Heterozygous Homozygous 22382244 AT4G10770 transporter F family member 3-like Uncharacterized protein oligopeptide transporter 7 Heterozygous Homozygous 22436720 AT5G51030 NAD(P)- binding Rossmann-fold superfamily protein Heterozygous Homozygous 22958724 AT1G52150 Homeobox- Heterozygous Homozygous 23079581 N/A leucine zipper family protein / lipid-binding START domain- containing protein Uncharacterized protein Heterozygous Homozygous 23140914 AT5G20885 RING/U-box superfamily protein Heterozygous Homozygous 23143585 AT5G20885 RING/U-box Heterozygous Homozygous 23985531 AT3G23760 Heterozygous Homozygous 24402010 AT3G48860 Heterozygous Homozygous 3969285 AT2G32230 superfamily protein transferring glycosyl group transferase coiled-coil protein proteinaceous RNase P 1 61 Table 6. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Kidd’s D-8’ compared to both the reference genome and ‘Autumn Gala’. Chromosome Gene 1 Mdg_01g018420 Variant Type Frameshift Variant location Gene ortholog 29044966 AT3G21790 10 Mdg_10g003900 Frameshift and start codon lost 4978943 AT4G15760 Gene Description UDP-Glycosyltransferase superfamily protein monooxygenase 1 Table 7. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Autumn Gala’ compared to both the reference genome and ‘Kidd’s D-8’. Chromosome Gene 1 Mdg_01g008670 Variant Type Stop codon gained and conservative in-frame insertion Variant location Gene ortholog 19075786 AT1G31280 Gene Description Argonaute family protein Table 8. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where, when called heterozygous for ‘Kidd’s D-8’, ‘Autumn Gala’ SNPs were called homozygous. Chromosome Gene Variant Type Kidd’s D-8 Genotype Autumn Gala Genotype Variant location Variant length Gene ortholog Gene Description 6 6 Mdg_06g010800 Deletion Heterozygous Homozygous 21907728- 21912511 4783 At1G64510 Translation elongation factor EF1B/ribosomal protein S6 family protein Mdg_06g011660 Deletion Heterozygous Homozygous 23072809- 23072911 102 At1G32928 Avr9/Cf-9 rapidly elicited protein Table 9. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Kidd’s D-8’ compared to both the reference genome and ‘Autumn Gala’. Chromosome 1 Gene Mdg_01g020600 Deletion Variant Type Variant location Variant length Gene ortholog 30865375- 30865454 AT5G16750 79 Gene Description Transducin family protein / WD-40 repeat family protein Table 10: Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Autumn Gala’ compared to both the reference genome and ‘Kidd’s D-8’. Chromosome 6 Gene Mdg_06g011460 Deletion Variant Type Variant location 22686399- 22686629 62 Variant length Gene ortholog 230 AT4G00870 Gene Description protein dimerization activity 3.3.3. DNA Variants in ‘September Wonder Fuji’ and ‘Aztec Fuji’ from Haploid Gene Alignments Variant calling for the ‘Fuji’ cultivars according to the ‘Gala’ haploid genome identified several candidate genes potentially responsible for the difference in development and maturity (Tables 11, 12, 13, 14, 15). There were 5 SNPs called heterozygous in ‘September Wonder Fuji’ and homozygous in ‘Aztec Fuji’ while 12 SNPs were called the converse: homozygous in ‘September Wonder Fuji’ and heterozygous and ‘Aztec Fuji’. 13 SNPs were called unique to ‘September Wonder Fuji’, while 8 unique SNPs were called for ‘Aztec Fuji’. There were no structural variants shared between ‘September Wonder Fuji’ and ‘Aztec Fuji’ near a gene. 28 structural variants were unique to ‘September Wonder Fuji’ and 54 structural variants unique to ‘Aztec Fuji’ were identified. 63 Table 11. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where, when called homozygous for ‘September Wonder Fuji’, ‘Aztec Fuji’ SNPs were called heterozygous, and vice versa. Chromosome Gene Variant Type September Wonder Fuji Genotype Aztec Fuji Genotype Variant location Gene ortholog Gene Description Mdg_01g004110 Frameshift Homozygous Heterozygous 11555142 AT2G13600 1 2 2 2 3 3 5 5 6 8 9 - - - 10 10 10 Mdg_02g006630 Mdg_02g023100 Mdg_02g023350 Mdg_03g000920 Mdg_03g005160 Mdg_05g002820 Mdg_05g014460 Mdg_06g005500 Mdg_08g011080 Mdg_09g016940 Mdg_scaffold227g00 0070 Mdg_scaffold333g00 0030 Mdg_scaffold644g00 0020 Mdg_10g003610 Mdg_10g007370 Mdg_10g010190 Homozygous Heterozygous 5194646 AT2G19540 Stop codon gained Homozygous Heterozygous Homozygous Homozygous Heterozygous Heterozygous Heterozygous Homozygous Stop codon lost and splice region variant Frameshift and splice region variant Frameshift and start codon lost Stop codon gained Stop codon gained Stop codon gained Frameshift Heterozygous Homozygous Heterozygous Splice donor, region, and intron variant Splice donor and intron variant Frameshift Heterozygous Homozygous Heterozygous Homozygous Heterozygous Heterozygous Homozygous Homozygous Homozygous 32196900 AT3G12490 32565350 AT3G14470 799574 AT1G69770 4691984 N/A 26958605 AT5G17680 7569463 AT4G02570 10412779 AT5G25610 16394615 AT1G17720 66088 N/A 4675007 AT1G06930 TPRXL Pentatricopeptide repeat (PPR) superfamily protein Transducin family protein / WD-40 repeat family protein cystatin B NB-ARC domain- containing disease resistance protein chromomethylase 3 Uncharacterized protein disease resistance protein (TIR-NBS- LRR class) cullin 1 BURP domain- containing protein Protein phosphatase 2A%2C regulatory subunit PR55 Uncharacterized protein Ribosomal protein L13 family protein disease resistance protein (TIR-NBS- LRR class) Frameshift Homozygous Heterozygous 31840 AT3G07110 Heterozygous 3048 AT1G69550 Homozygous Stop codon gained Frameshift Homozygous Heterozygous 4594552 AT1G22730 MA3 domain- Homozygous Heterozygous 9385826 AT2G21580 containing protein Ribosomal protein S25 family protein Heterozygous Homozygous 17068441 ATMG00860 DNA/RNA polymerases superfamily protein Stop codon gained Stop codon lost and splice region variant 64 Table 11 (cont’d) 10 10 10 11 11 11 11 11 12 13 14 14 14 14 14 15 Mdg_10g012560 Mdg_10g014210 Mdg_10g021190 Heterozygous Homozygous Stop codon lost and splice region variant Stop codon lost and splice region variant Frameshift Heterozygous Homozygous Heterozygous Homozygous 21409000 AT5G23590 23858541 AT1G43760 33158308 N/A DNAJ heat shock N-terminal domain-containing protein DNAse I-like superfamily protein protein FAR1- RELATED SEQUENCE 5-like [Prunus avium] Mdg_11g008880 Frameshift Homozygous Heterozygous 8134959 AT1G06740 MuDR family Mdg_11g009770 Mdg_11g010200 Mdg_11g015780 Mdg_11g024440 Mdg_12g017710 Mdg_13g001910 Mdg_14g001740 Mdg_14g003090 Mdg_14g004670 Mdg_14g005920 Mdg_14g012840 Mdg_15g028330 Heterozygous Homozygous Homozygous Stop codon lost and splice region variant Stop codon gained Splice acceptor and intron variant Frameshift Heterozygous Homozygous Heterozygous Homozygous Heterozygous 9051436 N/A 9536837 AT1G40087 19147535 N/A 36321824 N/A Heterozygous Homozygous 26630528 AT4G11150 Heterozygous Homozygous 1443149 AT1G23200 Homozygous Heterozygous 1721359 AT5G39340 Heterozygous Homozygous 2992565 AT2G19130 Homozygous Heterozygous 5126589 AT3G12010 Heterozygous Homozygous 6548252 N/A Heterozygous Homozygous 19579529 AT2G34320 Homozygous Heterozygous 31241492 N/A Splice donor and intron variant Stop codon gained Stop codon gained Stop codon gained Stop codon gained Splice acceptor and intron variant Stop codon gained Splice acceptor, splice region, and intron variant 65 transposase PREDICTED: metallothionein- like protein type 2 [Malus domestica] Plant transposase (Ptta/En/Spm family) Uncharacterized protein Uncharacterized protein vacuolar ATP synthase subunit E1 Plant invertase/pectin methylesterase inhibitor superfamily histidine- containing phosphotransmitter 3 S-locus lectin protein kinase family protein C18orf8 Uncharacterized protein Polynucleotidyl transferase%2C ribonuclease H- like superfamily protein Uncharacterized protein Table 11 (cont’d) 15 Mdg_15g031110 Frameshift Homozygous Heterozygous 40781377 ATMG00310 RNA-directed 15 15 17 Mdg_15g033100 Frameshift Homozygous Heterozygous 44562688 N/A Mdg_15g033790 Frameshift Heterozygous Homozygous 46091054 AT5G36930 Mdg_17g021670 Stop codon gained Homozygous Heterozygous 28156959 AT2G38995 DNA polymerase (reverse transcriptase)- related family protein Uncharacterized protein Disease resistance protein (TIR-NBS- LRR class) family O-acyltransferase (WSD1-like) family protein 66 Table 12. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘September Wonder Fuji’ compared to both the reference genome and ‘Aztec Fuji’. Chromosome Gene Variant Type Mdg_02g013000 Stop codon gained Variant location 11079927 Gene ortholog AT2G18570 Mdg_02g025220 Frameshift 34710788 AT5G19440 Mdg_03g010260 Mdg_03g016050 Mdg_07g001700 Mdg_scaffold422g000040 Stop codon gained Splice acceptor, region, missense, and intron variant Frameshift and missense Frameshift 10017807 23950438 N/A N/A 1802896 AT3G21640 22202 AT3G14470 Mdg_scaffold676g000050 Stop codon gained 22334 AT2G18280 Mdg_11g008910 Mdg_11g020960 Mdg_12g011930 Stop codon gained 8152436 N/A Uncharacterized protein Frameshift Frameshift 31148937 AT1G47490 RNA-binding protein 47C 18821575 AT3G54400 Gene Description UDP-Glycosyltransferase superfamily protein NAD(P)-binding Rossmann-fold superfamily protein Uncharacterized protein Uncharacterized protein FKBP-type peptidyl-prolyl cis- trans isomerase family protein NB-ARC domain-containing disease resistance protein tubby like protein 2 Eukaryotic aspartyl protease family protein kinase with tetratricopeptide repeat domain-containing protein Ribonuclease H-like superfamily protein ribonuclease Ps 2 2 3 3 7 - - 11 11 12 14 16 17 Mdg_14g017480 Frameshift 25768540 AT1G63500 Mdg_16g019850 Frameshift 21911304 AT4G29090 Mdg_17g024890 Frameshift 33085205 AT2G47300 67 Table 13. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Aztec Fuji’ compared to both the reference genome and ‘September Wonder Fuji’. Chromosome Gene Variant Type 2 4 8 9 - - 11 17 Mdg_02g017670 Mdg_04g017200 Mdg_08g020420 Mdg_09g014140 Mdg_scaffold395g000010 Mdg_scaffold824g000060 Mdg_11g025170 Mdg_17g019270 Stop codon gained Frameshift Frameshift Frameshift and splice region variant Splice acceptor and intron variant Splice acceptor, region, and intron variant Splice acceptor and intron variant Frameshift Variant location 19960266 Gene ortholog N/A Gene Description Uncharacterized protein 27362300 AT5G57580 Calmodulin-binding protein 28155987 28156026 N/A 12285390 N/A 3583 N/A 39547 AT4G08850 PREDICTED: mediator of RNA polymerase II transcription subunit 30 [Malus domestica] PREDICTED: replication protein A 70 kDa DNA-binding subunit B-like [Pyrus x bretschneideri] PREDICTED: Regulator of rDNA transcription protein 15 [Capsicum baccatum] Leucine-rich repeat receptor-like protein kinase family protein 37256975 N/A Uncharacterized protein 24704155 AT1G32900 UDP-Glycosyltransferase superfamily protein 68 Table 14. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘September Wonder Fuji’ compared to both the reference genome and ‘Aztec Fuji’. Chromosome Gene 1 Mdg_01g020600 Variant Type Variant location Deletion 30865375- 30865454 Variant length Gene ortholog 79 AT5G16750 2 2 2 2 2 2 3 3 9 9 9 9 9 10 10 10 10 10 Mdg_02g004740 Translocation 3683286- 2294687 0 AT4G13360 Mdg_02g004740 Deletion 3679515- 3679842 327 AT4G13360 Mdg_02g019670 Deletion 27419843- 27419991 148 AT1G66920 Mdg_02g019980 Deletion 28010781- 28011216 435 AT3G22690 Mdg_02g021650 Deletion 30385003- 30385403 400 AT4G12330 Mdg_02g024180 Deletion 33601979- 33602889 910 AT5G59100 Mdg_03g006990 Deletion 6573562- 6573667 105 AT1G61190 Mdg_03g010230 Deletion 9876470- 9878074 1604 AT3G09740 Mdg_09g001350 Deletion 846075- 850886 4811 AT2G03050 Mdg_09g004960 Deletion 3480511- 3481288 777 AT2G03340 Mdg_09g017070 Deletion 16643045- 16644019 974 AT1G73040 Mdg_09g017220 Deletion 16940019- 16940282 263 AT1G51850 Mdg_09g019280 Deletion 22653283- 22663204 9921 Mdg_10g026700 Duplication 38749505- 38749588 83 N/A N/A Mdg_10g000300 Deletion 529946- 533651 3705 AT1G56070 Mdg_10g008880 Deletion 12364570- 12364622 52 AT2G25970 Mdg_10g009130 Deletion 12828553- 12828700 147 AT1G49590 Mdg_10g013060 Deletion 22287395- 22290799 3404 N/A 69 Gene Description Transducin family protein / WD-40 repeat family protein ATP-dependent caseinolytic (Clp) protease/crotonase family protein ATP-dependent caseinolytic (Clp) protease/crotonase family protein Protein kinase superfamily protein LOW protein: PPR containing-like protein cytochrome P450%2C family 706%2C subfamily A%2C polypeptide 7 Subtilisin-like serine endopeptidase family protein LRR and NB-ARC domains-containing disease resistance protein syntaxin of plants 71 Mitochondrial transcription termination factor family protein WRKY DNA- binding protein 3 Mannose-binding lectin superfamily protein Leucine-rich repeat protein kinase family protein Uncharacterized protein PREDICTED: polyphenol oxidase I, chloroplastic-like [Malus domestica] Ribosomal protein S5/Elongation factor G/III/V family protein KH domain- containing protein C2H2 and C2HC zinc fingers superfamily protein Uncharacterized protein Table 14 (cont’d) 10 11 11 13 14 15 16 16 17 Mdg_10g023800 Deletion 36081008- 36081229 221 AT5G46250 Mdg_11g000460 Mdg_11g000470 Mdg_11g015610 Mdg_11g015620 Mdg_11g015630 Deletion 381246- 384184 2938 N/A Inversion 18524249- 18689300 165051 AT1G05750 AT2G01820 AT1G08440 Mdg_13g004400 Deletion 3370842- 3371477 635 AT5G51700 Mdg_14g013400 Deletion 20742477- 20744226 1749 AT1G13320 Mdg_15g028520 Deletion 32006359- 32011367 5008 AT4G38180 Mdg_16g014140 Deletion Mdg_16g015340 Deletion 12033369- 12034152 13850067- 13850220 783 153 AT5G49360 AT5G18640 Mdg_17g004250 Deletion 3098599- 3098656 57 AT3G28460 RNA-binding protein PREDICTED: F- box/kelch-repeat protein At3g06240- like [Prunus mume] Tetratricopeptide repeat (TPR)-like superfamily protein Leucine-rich repeat protein kinase family protein Aluminum activated malate transporter family protein cysteine and histidine-rich domain-containing protein RAR1 protein phosphatase 2A subunit A3 FAR1-related sequence 5 beta-xylosidase 1 alpha/beta- Hydrolases superfamily protein Methyltransferase 70 Table 15. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Aztec Fuji’ compared to both the reference genome and ‘September Wonder Fuji’. Chromosome Gene 1 Mdg_01g014680 Variant Type Variant location Deletion 25632047- 25632573 Variant length Gene ortholog 526 N/A 1 1 1 2 2 2 2 2 3 3 4 4 4 5 5 5 5 5 5 6 Mdg_01g017670 Deletion 28411734-28411789 55 AT5G23850 Mdg_01g017860 Deletion 28594343-28594557 214 AT1G64940 Mdg_01g019640 Deletion 30141518-30141733 215 AT5G53130 Mdg_02g005240 Deletion 4080234-4080296 199 AT2G25770 Mdg_02g009940 Deletion 7976438-7976601 163 AT4G03500 Mdg_02g020250 Deletion 28512421-28512581 160 AT3G52990 Mdg_02g026650 Deletion 36559368-36559426 58 N/A Mdg_02g026700 Deletion 36617991-36618103 112 AT1G05590 Mdg_03g010410 Deletion 10277516-10277616 100 AT2G28450 Mdg_03g012370 Deletion 14331372-14343997 12625 N/A Mdg_04g004750 Deletion 5081867-5081918 Mdg_04g005340 Mdg_04g020430 Deletion Deletion 5786040-5786131 30274880-30275765 51 91 885 AT3G52050 AT5G47090 N/A Mdg_05g016670 Deletion 29705370-29711520 6150 AT4G29035 Mdg_05g017420 Deletion 30528837 299 N/A Mdg_05g023310 Deletion 38382749-38391092 8343 AT5G45160 Mdg_05g030220 Deletion 44860570-44861560 990 N/A Mdg_05g030280 Deletion 44906222-44911226 5004 AT1G11410 Mdg_05g032370 Deletion 46450774-46450871 97 AT5G47540 Mdg_06g010730 Translocation 21831333-39066986 0 N/A 71 Gene Description Uncharacterized protein O- glucosyltransferase rumi-like protein (DUF821) cytochrome P450%2C family 87%2C subfamily A%2C polypeptide 6 cyclic nucleotide gated channel 1 Polyketide cyclase/dehydrase and lipid transport superfamily protein Ankyrin repeat family protein Pyruvate kinase family protein Uncharacterized protein beta- hexosaminidase 2 zinc finger (CCCH- type) family protein Uncharacterized protein 5'-3' exonuclease family protein coiled-coil protein PREDICTED: Ubiquitin-like domain-containing protein Plant self- incompatibility protein S1 family Uncharacterized protein Root hair defective 3 GTP-binding protein (RHD3) PREDICTED: G- type lectin S- receptor-like serine/threonine- protein kinase At1g11410 isoform X5 [Pyrus x bretschneideri] S-locus lectin protein kinase family protein Mo25 family protein Uncharacterized protein Table 15 (cont’d) 6 8 8 9 9 10 10 10 10 11 11 11 12 12 12 13 13 13 13 14 14 14 14 15 Mdg_06g010950 Deletion 22018714-22021699 2985 AT2G39020 Mdg_08g006030 Deletion 5046920-5048082 1162 AT5G26180 Mdg_08g006120 Mdg_08g006130 Deletion 5107082-5117560 10478 AT3G54920 AT5G12020 Mdg_09g012280 Translocation 9996582-19761301 0 AT3G22170 Mdg_09g006020 Deletion 4303090-4303090 100 AT3G29635 Mdg_10g011550 Translocation 19649141-29189379 0 AT3G07940 Mdg_10g006820 Deletion 8591448-8591622 174 AT5G60670 Mdg_10g013840 Deletion 23459141-23459484 343 AT2G17250 Mdg_10g016100 Deletion 26450192-26450371 179 AT5G48740 Mdg_11g020100 Translocation 29954863-35074315 0 AT4G08850 Mdg_11g011870 Deletion 11516225-11528837 12786 N/A Mdg_11g025090 Deletion Mdg_12g012240 Deletion 37165867-37173784 19214211-19215426 7917 1215 AT3G63270 N/A Mdg_12g012250 Deletion 19219111-19238373 19262 N/A Mdg_12g016070 Deletion 24495858-24503183 7325 AT3G27070 Mdg_13g012930 Deletion 10869601-10869986 385 AT5G49930 Mdg_13g014810 Deletion 13193422-13208669 15411 N/A Mdg_13g019160 Deletion Mdg_13g019830 Deletion 20442610-20442716 21599467-21599532 106 65 AT3G23920 N/A Mdg_14g012510 Translocation 19087046-37282952 0 AT5G42050 Mdg_14g018770 Translocation 26890519-23727181 0 N/A Mdg_14g002660 Deletion Mdg_14g018140 Deletion 2470299-2470512 26394672-26402408 213 7736 AT5G06140 N/A Mdg_15g001620 Deletion 1102955-1103055 100 AT4G16310 Acyl-CoA N- acyltransferases (NAT) superfamily protein S-adenosyl-L- methionine- dependent methyltransferases superfamily protein Pectin lyase-like superfamily protein 17.6 kDa class II heat shock protein far-red elongated hypocotyls 3 HXXXD-type acyl- transferase family protein Calcium-dependent ARF-type GTPase activating protein family Ribosomal protein L11 family protein CCAAT-binding factor Leucine-rich repeat protein kinase family protein Leucine-rich repeat receptor-like protein kinase family protein Uncharacterized protein Nuclease Uncharacterized protein Uncharacterized protein translocase outer membrane 20-1 zinc knuckle (CCHC-type) family protein Uncharacterized protein beta-amylase 1 PREDICTED: replication protein A 70 kDa DNA- binding subunit B [Prunus persica] DCD (Development and Cell Death) domain protein PREDICTED: cyclic dof factor 5- like [Malus domestica] sorting nexin 1 Uncharacterized protein LSD1-like 3 72 Table 15 (cont’d) 15 15 15 15 15 16 17 17 17 Mdg_15g025060 Deletion 24292486-24296967 4481 AT4G39950 Mdg_15g029100 Deletion 33545360-33545696 336 AT1G68470 Mdg_15g033320 Deletion 44992873-44993050 177 AT1G22275 Mdg_15g033690 Deletion 45819172-45827385 8213 AT1G11870 Mdg_15g039310 Deletion 54042100-54042164 64 AT1G59740 Mdg_16g019230 Mdg_16g019240 Deletion 20885340-20887554 2214 AT4G29090 Mdg_17g001370 Deletion 926481-927185 704 AT2G34930 Mdg_17g002720 Deletion 1838393-1839478 1085 AT5G04700 Mdg_17g019520 Deletion 25028238-25033213 4975 AT5G20230 cytochrome P450%2C family 79%2C subfamily B%2C polypeptide 2 Exostosin family protein Myosin heavy chain-related protein Seryl-tRNA synthetase Major facilitator superfamily protein Ribonuclease H- like superfamily protein disease resistance family protein / LRR family protein Ankyrin repeat family protein blue-copper- binding protein 3.3.4. DNA Variants in ‘Maslin’ and ‘Cripps Pink’ from Haploid Genome Alignments Variant calling of ‘Pink Lady’ cultivars mapped to the ‘Gala’ diploid reference genome resulted in many different candidate genes potentially causing faster development and earlier maturation in ‘Maslin’ than ‘Cripps Pink’ (Tables 16, 17, 18, 19, 20). Of the SNPs shared between both cultivars, 12 SNPs were called homozygous and heterozygous in ‘Maslin’ and ‘Cripps Pink’, respectively. 6 SNPs unique to ‘Maslin’ and 10 SNPs unique to ‘Cripps Pink’ were identified. A single structural variant was called that is shared by both cultivars, heterozygous in ‘Maslin’ and homozygous in ‘Cripps Pink’. We identified 34 structural variants unique to ‘Maslin’, and 12 structural variants unique to ‘Cripps Pink’. 73 Table 16. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where, when called homozygous for ‘Maslin’, ‘Cripps Pink’ SNPs were called heterozygous. Maslin Chromosome Gene Genotype Gene Description Variant Type Frameshift Homozygous Cripps Pink Genotype Heterozygous Variant location 6923933 Gene ortholog AT5G17880 Mdg_02g008760 2 2 2 3 4 5 9 10 13 13 13 17 disease resistance protein (TIR-NBS- LRR class) Heterozygous 12406291 AT5G66900 Disease resistance protein (CC-NBS- LRR class) family 29949183 AT3G14470 NB-ARC domain- containing disease resistance protein Mdg_02g013760 Frameshift Homozygous Heterozygous Mdg_02g021330 Mdg_03g006990 Homozygous Splice acceptor, splice region, and intron variant Frameshift Homozygous Heterozygous 6574422 AT1G61190 LRR and NB-ARC Mdg_04g012150 Frameshift Homozygous Heterozygous 21218517 N/A Mdg_05g002920 Mdg_09g002180 Mdg_10g014740 Mdg_13g015550 Stop codon gained Splice donor and intron variant Stop codon gained Stop codon gained Homozygous Heterozygous 4792319 AT5G58430 Homozygous Heterozygous 1402559 AT5G40510 Homozygous Heterozygous 24617859 N/A Homozygous Heterozygous 14141799 AT3G10310 Mdg_13g017980 Frameshift Homozygous Heterozygous 18344552 AT5G43580 domains- containing disease resistance protein Uncharacterized protein exocyst subunit exo70 family protein B1 Sucrase/ferredoxin- like family protein PREDICTED: CDT1-like protein a, chloroplastic [Pyrus x bretschneideri] P-loop nucleoside triphosphate hydrolases superfamily protein with CH (Calponin Homology) domain-containing protein Serine protease inhibitor%2C potato inhibitor I- type family protein 20225335 AT1G10000 Ribonuclease H- like superfamily protein Uncharacterized protein 4277370 N/A Mdg_13g018970 Mdg_17g005610 Stop codon gained Stop codon gained Homozygous Heterozygous Homozygous Heterozygous 74 Table 17. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Maslin’ compared to both the reference genome and ‘Cripps Pink’. Chromosome Gene Variant Type 2 8 10 12 17 - Mdg_02g010040 Frameshift Mdg_08g016540 Mdg_10g015210 Mdg_12g015750 Mdg_17g002760 Mdg_scaffold713g000010 Stop codon gained and splice region variant Frameshift Splice donor and intron variant Stop codon lost and splice region variant Frameshift Variant location 8088886 Gene ortholog N/A 22299333 AT4G38180 Gene Description PREDICTED: ubiquitin domain- containing protein DSK2b-like isoform X1 [Prunus mume] FAR1-related sequence 5 25228705 N/A Uncharacterized protein 24192220 AT3G51830 1866004 AT5G04700 SAC domain-containing protein 8 Ankyrin repeat family protein 11768 AT5G05800 Myb/SANT-like DNA-binding domain protein Table 18. SNPs located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Cripps Pink’ compared to both the reference genome and ‘Maslin’. Chromosome Gene Mdg_02g020820 Mdg_05g023530 Mdg_07g004890 Mdg_08g016450 Mdg_10g012140 Variant Type Frameshift and missense Frameshift Variant location 29381235 Gene ortholog AT5G38260 38623658 N/A Stop codon gained 5246684 AT3G14470 Frameshift 22182455 N/A Start codon lost 20724755 AT2G02650 Mdg_10g027910 Frameshift 40265746 N/A Mdg_12g002000 Stop codon gained 2212591 AT1G50830 Mdg_12g015590 Mdg_14g013960 Stop codon gained and splice region variant Frameshift 23985583 N/A 21590378 AT4G23160 Mdg_scaffold197g000010 Stop codon lost and aplice region variant 6145 N/A Gene Description Protein kinase superfamily protein Uncharacterized protein NB-ARC domain-containing disease resistance protein Uncharacterized protein Ribonuclease H-like superfamily protein PREDICTED: RINT1-like protein MAG2 [Malus domestica] Aminotransferase-like%2C plant mobile domain family protein Uncharacterized protein cysteine-rich RECEPTOR-like kinase Uncharacterized protein 2 5 7 8 10 10 12 12 14 - Table 19. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where, when called homozygous for ‘Maslin’, ‘Cripps Pink’ SNPs were called heterozygous and vice versa. Chromosome Gene Variant Type Maslin Genotype Cripps Pink Genotype 17 Mdg_17g007020 Deletion Heterozygous Homozygous Variant location 5411012- 5411211 Variant length 199 Gene Gene ortholog Description AT5G15680 ARM repeat superfamily protein 75 Table 20. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Maslin’ compared to both the reference genome and ‘Cripps Pink’. Chromosome Gene 2 2 2 2 3 5 5 6 6 9 9 9 9 10 10 10 10 11 11 11 11 12 12 12 12 12 12 13 13 14 Mdg_02g003910 Mdg_02g004330 Mdg_02g006520 Mdg_02g017160 Mdg_03g013050 Mdg_05g000290 Deletion Deletion Deletion Variant Type Variant location 2942895- 2943260 3238270- 3238737 5139871- 5140431 18570329- 18576272 16134306- 16136604 870334-878413 Deletion Deletion Deletion Mdg_05g031220 Deletion Mdg_06g019700 Deletion Mdg_06g020940 Deletion Mdg_09g007580 Deletion Mdg_09g007720 Deletion Mdg_09g013250 Deletion Mdg_09g017070 Deletion Mdg_10g013060 Deletion Mdg_10g015720 Deletion Mdg_10g024300 Deletion Mdg_10g027920 Deletion Mdg_11g008630 Mdg_11g002210 Translocation Deletion Mdg_11g002400 Deletion Mdg_11g002950 Deletion Mdg_12g001820 Deletion Mdg_12g002650 Deletion Mdg_12g007620 Deletion Mdg_12g008920 Deletion Mdg_12g013300 Deletion Mdg_12g015080 Deletion Mdg_13g017160 Deletion Mdg_13g020580 Deletion Mdg_14g011280 Deletion 45612762- 45633696 31681850- 31692400 32910050- 32923020 5802137- 5802532 5924767- 5924924 11054344- 11055262 16643049- 16644019 22286696- 22291159 25960991- 25961245 36503422- 36503579 40282120- 40285495 7856923 1994363- 2000908 2157189- 2158297 2693145- 2693209 1999892- 2000850 2862700- 2863349 9098264- 9098585 11202669- 11205088 20963025- 20963078 23455300- 23456238 16852131- 16852345 23404689- 23404821 16784501- 16793536 76 Variant length 365 Gene ortholog AT5G36930 AT3G24503 Gene Description Disease resistance protein (TIR- NBS-LRR class) family aldehyde dehydrogenase 2C4 AT5G11800 K+ efflux antiporter 6 AT1G01950 armadillo repeat kinesin 2 AT3G05850 MuDR family transposase AT4G10300 20934 AT1G66250 10550 AT1G67000 12970 AT1G21280 AT3G18670 RmlC-like cupins superfamily protein O-Glycosyl hydrolases family 17 protein Protein kinase superfamily protein Copia-like polyprotein/retrotransposon Ankyrin repeat family protein AT1G03620 ELMO/CED-12 family protein AT1G06520 AT1G73040 4463 N/A 254 AT4G13780 AT4G18040 AT4G23180 AT2G29040 AT4G23160 AT5G60900 AT5G55850 N/A glycerol-3-phosphate acyltransferase 1 Mannose-binding lectin superfamily protein Uncharacterized protein methionine-tRNA ligase%2C putative / methionyl-tRNA synthetase%2C putative / MetRS eukaryotic translation initiation factor 4E cysteine-rich RLK (RECEPTOR- like protein kinase) 10 Exostosin family protein cysteine-rich RECEPTOR-like kinase receptor-like protein kinase 1 RPM1-interacting protein 4 (RIN4) family protein Uncharacterized protein AT3G18830 AT1G67810 polyol/monosaccharide transporter 5 sulfur E2 2419 N/A Uncharacterized protein AT2G17080 hypothetical protein (DUF241) AT5G23960 terpene synthase 21 AT5G43860 chlorophyllase 2 AT5G58430 9035 AT1G64040 exocyst subunit exo70 family protein B1 type one serine/threonine protein phosphatase 3 467 560 5943 2298 8079 395 157 918 970 157 3375 0 6545 1108 64 958 649 321 53 938 214 132 Table 20 (cont’d) 15 15 15 15 Mdg_15g011940 Deletion Mdg_15g025510 Deletion Mdg_15g038710 Deletion Mdg_15g039130 Deletion 9287941- 9288435 25067107- 25068788 53365293- 53365878 53856135- 53856209 494 AT4G31980 1681 AT4G37030 585 AT2G40890 74 AT5G42260 PPPDE thiol peptidase family protein membrane protein cytochrome P450%2C family 98%2C subfamily A%2C polypeptide 3 beta glucosidase 12 77 Table 21. Structural variants located within genes mapped to the ‘Gala’ haploid reference genome where all variants are unique to ‘Cripps Pink’ compared to both the reference genome and ‘Maslin’. Chromosome Gene 4 Mdg_04g003020 Variant Type Variant location Variant length Gene ortholog Deletion AT1G05785 1141 3394853- 3395994 5 5 8 9 9 11 12 14 14 15 17 Mdg_05g013960 Deletion Mdg_05g023470 Mdg_05g023480 Deletion Mdg_08g010660 Mdg_08g010670 Mdg_09g013060 Mdg_09g013070 Inversion Deletion Mdg_09g015190 Deletion Mdg_11g002670 Deletion Mdg_12g012270 Deletion Mdg_14g005750 Deletion Mdg_14g020930 Deletion Mdg_15g015240 Deletion Mdg_17g020190 Deletion 26255452- 26255558 38568224- 38574419 9857204- 9876445 10859222- 10861376 13751408- 13755530 2380774- 2380844 19263940- 19267692 6381972- 6382893 28736671- 28742421 12281144- 12281195 26196912- 26197057 106 6195 AT2G17350 AT1G56130 AT4G02550 19241 2154 AT3G63520 AT1G55000 4122 70 3752 921 5750 51 145 AT4G37030 AT3G20870 AT3G10180 AT3G10300 AT3G18290 AT5G35080 AT5G36110 Gene Description Got1/Sft2-like vescicle transport protein family beta- mannosyltransferase- like protein Leucine-rich repeat transmembrane protein kinase Myb/SANT-like DNA-binding domain protein carotenoid cleavage dioxygenase 1 peptidoglycan- binding LysM domain-containing protein membrane protein ZIP metal ion transporter family P-loop containing nucleoside triphosphate hydrolases superfamily protein Calcium-binding EF- hand family protein zinc finger protein- like protein ER lectin-like protein cytochrome P450%2C family 716%2C subfamily A%2C polypeptide 1 3.3.4. Preparation for Transcriptomic Study A comprehensive transcriptome study was designed to complement our DNA variant analyses. Fruit tissue from all six cultivars was collected throughout the 2023 growing season between May 23 and harvest date of each cultivar. Collections from May 23 to July 11 were of whole fruit, while July 18 – harvest date, hand dissections were done to isolate peel, cortex, core, and seed tissues. Table 22 indicates collection details for timing and type of collection of tissue. Our analysis of fruit growth between sport and standard maturity cultivars, as described in chapter 2, suggests that fruit in early maturing lines for all three varieties increased size at a 78 faster rate early in development than the late-maturing lines. This suggests that the causative mutation(s) for the altered maturity time for each of our varieties likely impacts a gene involved in early fruit development, or carpel development preceding it. For this reason, upcoming transcriptome studies should be done using RNA extracted from fruit harvested at these dates between 23 May and 26 September 2023 (Table 22; and the Fig. 21 below). 79 Table 22. Tissue collections for RNA extraction and sequencing, including date, days after full bloom (DAFB), and type of tissue. DAFB Julian Date Tissue Date 5/23/2023 6/6/2023 6/13/2023 6/20/2023 7/3/2023 7/11/2023 7/18/2023 7/25/2023 8/1/2023 14 28 35 42 55 63 70 77 84 9/13/2023 127 Whole fruit, cut in half Whole fruit, cut in quarters Whole fruit, cut in quarters Whole fruit, cut in 1/16’s Whole fruit, cut many pieces Whole fruit, cut many pieces Peel, cortex, seed Peel, cortex, core, seed Peel, cortex, core, seed Peel, cortex, core, seed 143 157 164 171 184 192 199 206 213 256 80 Figure 20. Volumetric growth curve of standard harvesting ‘Kidd’s D-8’ and its late-harvesting sport, ‘Autumn Gala’ according to growing degree hours (GDH). The 6 first blue lines indicate dates (Table 22) when whole fruit tissue was collected for RNA extraction and sequencing. These dates refer to fruit at the cell division (23 May), cell division and cell expansion overlap (6, 13, 20 June), and cell expansion stages of development. The last 4 black lines indicate dates (Table 22) when dissected fruit tissue was collected for RNA extraction and sequencing (Fig. 21). 81 Figure 21. Example of fruit tissue dissections that were collected 25 July 2023 and later. 3.4. DISCUSSION 3.4.1. Preliminary Genomic Analysis Our preliminary genomic results indicate many mechanisms that need to be assessed for causation of early or late maturation. Our phenotypical data (chapter 2) suggests that the genetic differences between a maturity sport and its progenitor, or a standard maturity cultivar of the same variety, may relate to genes that are associated with early fruit development. Therefore, variants in genes associated with cell division would be strong candidates for causative mutations for the maturity phenotypes, unlike one of the few genes listed above potentially related to photosynthesis (Table 20). In ‘Gala’, we find the least number of variants. This is 82 likely due to the close relation to the reference genome ‘Gala Haploid’, which was derived from budwood from the original 'Gala' accession a.k.a. 'Kidd's D-8 Gala' (Sun et al., 2020). Even if there are other homologs in the genome encoding proteins that maintain the same or similar functions as gene ‘Mdg_04A010840’ (which encodes a motor family protein), such genetic events may cause a reduction of total expression of the other genes and may result in lower rates of, as an example, cell division early in fruit development. Slower cell division early in fruit development may result in a ‘compound interest’ effect where slower cell division early in fruit development may significantly impact the rest of the fruit’s development. In ‘Fuji’ and ‘Pink Lady’ comparisons, we find many variants of genes encoding important proteins. These proteins are part of large families, and the causal genes regulating fruit maturation may be a result of many genes in a particular locus in the genome being differentially expressed. Transcriptomic analyses may provide more insight into the mechanisms behind final fruit maturation date by closing in on specific genes that have specific functions involved in contributing to the early or late maturity phenotype. Analysis of differential gene expression at different timepoints could also prove useful. Finding unexpressed genes that are in our variant lists may also assist in filtering out candidate genes, while also evaluating the pathways in which those genes are involved. Although we were not able to directly detect a large deletion of 2.8 Mb in ‘Autumn Gala’ with our short-read sequencing, our haploid results of 11 homozygous SNPs and 2 homozygous deletions within a 2.8 Mb region in the ‘Gala’ haploid genome agree with the 2.8 Mb deletion found in ‘Autumn Gala’ by Ban et al. (2022). Since all ‘Autumn Gala’ SNP variants are homozygous in the matching 2.8 Mb region, we can infer that we are likely also seeing a 2.8 Mb deletion in ‘Autumn Gala’. It is unclear from our results, however, if the candidate Mdact7 gene 83 ‘MD06G1127300’ that Ban et al. (2022) proposed as the primary candidate for the delayed maturation in ‘Autumn Gala’. What we did find was other homologs of gene ‘MD06G1127300’ [from the ‘Golden Delicious’ genome (Daccord et al., 2017)] in the ‘Gala’ diploid and haploid genomes (X. Sun et al., 2020) responsible for actin-7 production. Plants often have multiple genes encoding for similar proteins critical to plant processes. It is possible that removing functionality of one single gene may not necessarily turn a plant process off, but rather delay the process. In the case of early maturing bud sports, there may be transcription factors enabled by mutation that promote more genetic expression of basic molecular processes, leading to the advancement of maturity. The phenotype of change in maturation could be due to a mechanism such as a mutational change in a particular protein encoded by one gene that is similar amongst all apple cultivars. Since harvest date is potentially such a complex trait, there may be many possible mechanisms that contribute to early- or late-maturation of an apple bud sport. It is possible that a slight manipulation of a particular gene involved in basic cell activity such as division early in fruit development could create a compound interest effect for the rest of the season, not just in size, but in developmental rate. Based on our findings in all 3 apple variety comparisons, since each earlier harvesting cultivar shows almost an immediate increase in rate of development after flowering, mechanisms that result in shifting harvest date in these three varieties commence early in fruit development. In the case of the non-functional actin-7 encoding gene proposed to confer late a maturity date by Ban et al. (2022), the suggestion is that a mutation in a gene providing basic cellular functions such as suppressed actin production may lead to a delay in maturity. It is unclear, however, how such a loss in gene function would not also result in a change in the canopy of the sport tree. 84 CHAPTER 4. Conclusions and Future Direction 4.1. INTRODUCTION In early apple fruit development, there are many mechanisms underlying crucial processes integrating the effects of both the whole tree as an organism and its environment. Fortunately, gene expression during early fruit development has been studied since the early 2000’s (Eccher et al., 2014). Eccher et al. (2014) noted in their concluding statements the knowledge gap in the study and knowledge of apple early fruit development and singled out possible interactions between the seeds and the cortex as an example. However, apple fruit are often seedless and no shift in maturation of seedless fruit to seeded fruit has, to our knowledge, been described. They also note the lack of information regarding controls, modulators, and molecular mechanisms that determine maturation rate and timing. They ask questions such as: “What kind of “molecular” competences does the apple fruit acquire during the maturation phase?”, “Which are the modulators of this process?”, and “Can this process be tuned by exogenous treatments?”. They mention how much work needs to be done and specifically note the need for “omics” work combined with classical physiological approaches that actually evaluate the tree in the orchard (Eccher et al., 2014). Thus, the complexity of fruit development and its interaction with fruit maturation remains to be further explored. Comprehensive analyses of mechanisms responsible for early or delayed maturation remain to be confirmed. The loss-of-function actin allele in the Ban et al. (2022) study remains to be expressed in apple tissue. The closer we as researchers come to uncovering these mechanisms, the better we may understand the complexities of the relationship between fruit development and fruit maturation. 85 4.2. DISCUSSION AND FUTURE DIRECTION 4.2.1. Discussion and Future Direction What our current findings suggest is that early fruit development and fruit maturation are interconnected, meaning that fruit size, harvest date, and harvest traits such as firmness, sugars, flavor, and internal ethylene production are, to an extent, predetermined or influenced by events early on in fruit development (Chapter 2 Results). The exciting benefit of our physiological data is that we now have a much more targeted approach to interrogate the genome and transcriptome. We have also improved how we look. The transcriptomic data may show gene expression differences that aren’t detectable by genomic analysis. The absence in variants of orthologs of the ACT7 gene in ‘Gala’ may be further revealed by gene expression differences between the two cultivars. The early emergence of phenotypic differences in growth rate between the bud sport and the control lines suggests the physiological processes leading to an early or late harvest date may also emerge very early in fruit development. If so, the early or delayed maturation date is very likely not strictly a function of ripening-related processes, but rather is derived from a season-long shift in metabolic activity. We now believe that we can look at transcriptomic events very early in development and link those to maturation rate/harvest date. Future endeavors should involve comprehensive physiological and molecular characterization of fruit tissue during cell differentiation, division, and expansion periods, as well as the rate at which these processes occur. This includes evaluating floral development, with a focus on late- stage ovary development, as well as investigating anatomical and molecular changes happening right after fertilization. Another analysis that should be performed is an evaluation of anatomical differences (e.g. cell size and number) as well as molecular differences in expanding the transcriptomic studies to flowers and flower buds before fertilization. Our preliminary results of 86 filtered variants in our ‘Gala’ comparison may show promise, but much work needs to be done to determine if one or more of the candidate genes causes the early or late maturity phenotype in apple bud sports. 4.2.2. Advantages of the Study An advantage of our study is that our ‘Gala’ comparison used the original germplasm. We were graciously given scions from the original ‘Autumn Gala’ limb sport, as well as the original progenitor tree. Over time, apples accumulate mutations and may slightly alter their expression of certain fruit phenotypes such as color and size, so this approach minimizes that risk. This original germplasm partially eliminates much of that variability. In our ‘Fuji’ comparison, we have the bud sport ‘September Wonder Fuji’, from the early ‘Fuji’ bud sport ‘Yataka Fuji’. We are comparing this early sport to a quite unrelated (relative to direct sport mutations) cultivar ‘Aztec Fuji’, for which there is no patent to the author’s knowledge. The separation in genetics of these two sports may give unique perspective to our study, especially if we find a similar mechanism in this ‘Fuji’ comparison as we may find in our ‘Gala’ or ‘Cripps Pink’ analyses. Our ‘Cripps Pink’ analysis between ‘Maslin Cripps Pink’ and its progenitor ‘Cripps Pink’ is also unique in that although ‘Maslin’ is a direct sport of ‘Cripps Pink’, we do not have the original germplasm from the original sport limb and progenitor tree. Through several generations of grafting, possible mutational change may have accumulated and could complicate the filtering process in finding a single region in which a candidate gene has been manipulated. Variation in somatic mutations even across individual branches of a single tree elaborated by Sun et al. (2023) underlies the importance of utilizing original tissue of progenitor and mutant. This may also be a reason why there are so few variants in our ‘Gala’ filtered variant results compared to the ‘Fuji’ and ‘Cripps Pink’ lines. It is certainly valuable to study these three separate 87 comparisons concurrently because of replication of maturity sport studies in apple, as well as the fact that these three commercial varieties are widely cultivated across the globe. Understanding these mechanisms will inevitably lead to greater advances in breeding efforts to establish unique cultivars with better and more desirable traits better tailored to growing method and region. 88 LITERATURE CITED AgroFresh. (n.d.). HarvistaTM near-harvest treatment from AgroFresh. AgroFresh. Retrieved November 27, 2023, from https://www.agrofresh.com/solutions/harvista/ Argenta, L. C., Wood, R. M., De Angelis Monteiro Terra, F., & Neuwald, D. A. (2022). Effect of preharvest ethylene inhibitor application on ‘Fuji’ apple on-tree maturation and quality after storage. Acta Horticulturae, 1344, 219–226. https://doi.org/10.17660/ActaHortic.2022.1344.32 Ban, S., El-Sharkawy, I., Zhao, J., Fei, Z., & Xu, K. (2022). An apple somatic mutation of delayed fruit maturation date is primarily caused by a retrotransposon insertion-associated large deletion. The Plant Journal, 111(6), 1609–1625. https://doi.org/10.1111/tpj.15911 Bergh, O. (1990). Effect of temperature during the first 42 days following full bloom on apple fruit growth and size at harvest. South African Journal of Plant and Soil, 7(1), 11–18. https://doi.org/10.1080/02571862.1990.10634530 Bhusal, N., Han, S.-G., & Yoon, T.-M. (2019). Impact of drought stress on photosynthetic response, leaf water potential, and stem sap flow in two cultivars of bi-leader apple trees (Malus × domestica Borkh.). Scientia Horticulturae, 246, 535–543. https://doi.org/10.1016/j.scienta.2018.11.021 Biale, J. B. (1964). Growth, Maturation, and Senescence in Fruits. Science, 146(3646), 880– 888. Blanpied, G. D., & Silsby, K. J. (1992). Predicting Harvest Date Windows for Apples. Cornell Cooperative Extension. https://hdl.handle.net/1813/3299 Bramlage, W. J., Greene, D. W., Autio, W. R., & McLaughlin, J. M. (1980). Effects of Aminoethoxyvinylglycine on Internal Ethylene Concentrations and Storage of Apples1. Journal of the American Society for Horticultural Science, 105(6), 847–851. https://doi.org/10.21273/JASHS.105.6.847 Brookfield, P., Murphy, P., Harker, R., & MacRae, E. (1997). Starch degradation and starch pattern indices; interpretation and relationship to maturity. Postharvest Biology and Technology, 11(1), 23–30. https://doi.org/10.1016/S0925-5214(97)01416-6 Chagné, D., Dayatilake, D., Diack, R., Oliver, M., Ireland, H., Watson, A., Gardiner, S. E., Johnston, J. W., Schaffer, R. J., & Tustin, S. (2014). Genetic and environmental control of fruit maturation, dry matter and firmness in apple (Malus × domestica Borkh.). Horticulture Research, 1, 14046. https://doi.org/10.1038/hortres.2014.46 Chassagne-Berces, S., Poirier, C., Devaux, M.-F., Fonseca, F., Lahaye, M., Pigorini, G., Girault, C., Marin, M., & Guillon, F. (2009). Changes in texture, cellular structure and cell wall composition in apple tissue as a result of freezing. Food Research International, 42(7), 788– 797. https://doi.org/10.1016/j.foodres.2009.03.001 Cho, H. J., Kim, G. H., & Choi, C. (2020). Differential gene expression and epigenetic 89 analyses between striped and blushed skinned sports of ‘Fuji’ apple. Scientia Horticulturae, 261, 108944. https://doi.org/10.1016/j.scienta.2019.108944 Chun, I.-J., Fallahi, E., Shafii, B., Tripepi, R. R., & Colt, W. M. (2002). Influence of rootstocks and microsprinkler fertigation on photosynthesis of “Fuji” apple trees. Journal of the American Pomological Society, 56, 23–29. Contreras, C., Alsmairat, N., & Beaudry, R. (2014). Prestorage Conditioning and Diphenylamine Improve Resistance to Controlled-atmosphere-related Injury in ‘Honeycrisp’ Apples. HortScience, 49(1), 76–81. https://doi.org/10.21273/HORTSCI.49.1.76 Daccord, N., Celton, J.-M., Linsmith, G., Becker, C., Choisne, N., Schijlen, E., van de Geest, H., Bianco, L., Micheletti, D., Velasco, R., Di Pierro, E. A., Gouzy, J., Rees, D. J. G., Guérif, P., Muranty, H., Durel, C.-E., Laurens, F., Lespinasse, Y., Gaillard, S., … Bucher, E. (2017). High- quality de novo assembly of the apple genome and methylome dynamics of early fruit development. Nature Genetics, 49(7), Article 7. https://doi.org/10.1038/ng.3886 Darwin, C. (1868). The Variation of Animals and Plants Under Domestication. J. Murray. DeJong, T. (2022). Concepts for understanding fruit trees. https://www.cabidigitallibrary.org/doi/epdf/10.1079/9781800620865.0000 Dong, Q. L., Yan, Z. Y., Liu, Z., & Yao, Y. X. (2011). Early ripening events caused by bud mutation in Beni Shogun apple. Russian Journal of Plant Physiology, 58(3), 439–447. https://doi.org/10.1134/S1021443711030034 Eccher, G., Ferrero, S., Populin, F., Colombo, L., & Botton, A. (2014). Apple (Malus domestica L. Borkh) as an emerging model for fruit development. Plant Biosystems - An International Journal Dealing with All Aspects of Plant Biology, 148(1), 157–168. https://doi.org/10.1080/11263504.2013.870254 Engelsma, J., & Ott, E. (2024). Cost of PGR’s in Michigan Apple Industry (A. Engelsma, Interviewer) [Personal communication]. Fallahi, E., Chun, I.-J., Neilsen, G. H., & Colt, W. M. (2001). Effects of Three Rootstocks on Photosynthesis, Leaf Mineral Nutrition, and Vegetative Growth of “Bc-2 Fuji” Apple Trees. Journal of Plant Nutrition, 24(6), 827–834. https://doi.org/10.1081/PLN-100103776 Flore, J. A., & Lakso, A. N. (1989). Environmental and Physiological Regulation of Photosynthesis in Fruit Crops. In Horticultural Reviews (pp. 111–157). John Wiley & Sons, Ltd. https://doi.org/10.1002/9781118060841.ch4 Foster, T. M., & Aranzana, M. J. (2018). Attention sports fans! The far-reaching contributions of bud sport mutants to horticulture and plant biology. Horticulture Research, 5(1), 44. https://doi.org/10.1038/s41438-018-0062-x Fujii, J., & Kennedy, R. (1984). Seasonal Changes in the Photosynthetic Rate in Apple Trees: A comparison between Fruiting and Nonfruiting Trees. American Society of Plant Biologists, 90 78(3), 519–524. Giné-Bordonaba, J., Echeverria, G., Duaigües, E., Bobo, G., & Larrigaudière, C. (2019). A comprehensive study on the main physiological and biochemical changes occurring during growth and on-tree ripening of two apple varieties with different postharvest behaviour. Plant Physiology and Biochemistry, 135, 601–610. https://doi.org/10.1016/j.plaphy.2018.10.035 Goffinet, M. C., Robinson, T. L., & Lakso, A. N. (1995). A comparison of ‘Empire’ apple fruit size and anatomy in unthinned and hand-thinned trees. Journal of Horticultural Science. https://www.tandfonline.com/doi/abs/10.1080/14620316.1995.11515307 Grappadelli, L. C., Lakso, A. N., & Flore, J. A. (1994). Early Season Patterns of Carbohydrate Partitioning in Exposed and Shaded Apple Branches. Journal of the American Society for Horticultural Science, 119(3), 596–603. https://doi.org/10.21273/JASHS.119.3.596 Griffith, C. (2022). Effects of Plant Growth Regulators and Weather on Bitter Pit Incidence in ‘Honeycrisp’ Apple [M.S., Michigan State University]. https://www.proquest.com/docview/2708306894/abstract/29EB20B06AAB4112PQ/1 Gussman, C. D., Goffreda, J. C., & Gianfagna, T. J. (1993). Ethylene Production and Fruit- softening Rates in Several Apple Fruit Ripening Variants. HortScience, 28(2), 135–137. https://doi.org/10.21273/HORTSCI.28.2.135 Harker, F. R., Kupferman, E. M., Marin, A. B., Gunson, F. A., & Triggs, C. M. (2008). Eating quality standards for apples based on consumer preferences. Postharvest Biology and Technology, 50(1), 70–78. https://doi.org/10.1016/j.postharvbio.2008.03.020 Iglesias, I., Echeverría, G., & Lopez, M. L. (2012). Fruit color development, anthocyanin content, standard quality, volatile compound emissions and consumer acceptability of several ‘Fuji’ apple strains. Scientia Horticulturae, 137, 138–147. https://doi.org/10.1016/j.scienta.2012.01.029 Isard, S., & Schaetz, R. (1998). EFFECTS OF WINTER WEATHER CONDITIONS ON SOIL FREEZING IN SOUTHERN MICHIGAN. Physical Geography, 19(1), 71–94. Kappel, F., & Flore, J. A. (1983). Effect of Shade on Photosynthesis, Specific Leaf Weight, Leaf Chlorophyll Content, and Morphology of Young Peach Trees. Journal of the American Society for Horticultural Science, 108(4), 541–544. https://doi.org/10.21273/JASHS.108.4.541 Khan, A., Carey, S. B., Serrano, A., Zhang, H., Hargarten, H., Hale, H., Harkess, A., Honaas, L., Carey, S. B., Serrano, A., Zhang, H., Hargarten, H., Hale, H., Harkess, A., & Honaas, L. (2022). A phased, chromosome-scale genome of ‘Honeycrisp’ apple (Malus domestica). Gigabyte, 2022, 1–15. https://doi.org/10.46471/gigabyte.69 Kim, Y. J., Ban, S., Cho, H. J., Han, A. R., & Choi, C. (2023). Comparative Analysis of Gene Expression between Early Maturation Mutant ‘Beni Shogun’ and ‘Fuji’ Cultivars during FruitDevelopment and Ripening. Horticulturae, 9(4), Article 4. https://doi.org/10.3390/horticulturae9040430 91 Kunihisa, M., Moriya, S., Abe, K., Okada, K., Haji, T., Hayashi, T., Kim, H., Nishitani, C., Terakami, S., & Yamamoto, T. (2014). Identification of QTLs for fruit quality traits in Japanese apples: QTLs for early ripening are tightly related to preharvest fruit drop. Breeding Science, 64(3), 240–251. https://doi.org/10.1270/jsbbs.64.240 Lakso, A., & Goffinet, M. (2017). Advances in understanding apple fruit development. In Achieving sustainable cultivation of apples. Burleigh Dodds Science Publishing Limited. Lakso, A. N. (1979). Seasonal Changes in Stomatal Response to Leaf Water Potential in Apple1. Journal of the American Society for Horticultural Science, 104(1), 58–60. https://doi.org/10.21273/JASHS.104.1.58 Lakso, A. N. (1983). Morphological and Physiological Adaptations for Maintaining Photosynthesis under Water Stress in Apple Trees. In R. Marcelle, H. Clijsters, & M. van Poucke (Eds.), Effects of Stress on Photosynthesis: Proceedings of a conference held at the ‘Limburgs Universitair Centrum’ Diepenbeek, Belgium, 22–27 August 1982 (pp. 85–93). Springer Netherlands. https://doi.org/10.1007/978-94-009-6813-4_8 Lakso, A. N. (2011). EARLY FRUIT GROWTH AND DROP - THE ROLE OF CARBON BALANCE IN THE APPLE TREE. Acta Horticulturae, 903, 733–742. https://doi.org/10.17660/ActaHortic.2011.903.102 Lakso, A. N. (2022). Challenges to interpreting internal and external factors limiting apple fruit growth, set and abscission. Acta Horticulturae, 1342, 317–328. https://doi.org/10.17660/ActaHortic.2022.1342.45 Lakso, A. N., White, M. D., & Tustin, D. S. (2001). SIMULATION MODELING OF THE EFFECTS OF SHORT AND LONG-TERM CLIMATIC VARIATIONS ON CARBON BALANCE OF APPLE TREES. Acta Horticulturae, 557, 473–480. https://doi.org/10.17660/ActaHortic.2001.557.63 Larson, J., & Kon, T. (2021). Apple Fruitlet Abscission Mechanisms. Horticultural Reviews, 49, 243–274. Lenth, R. V., Bolker, B., Buerkner, P., Giné-Vázquez, I., Herve, M., Jung, M., Love, J., Miguez, F., Piaskowski, J., Riebl, H., & Singmann, H. (2024). emmeans: Estimated Marginal Means, aka Least-Squares Means (1.10.2) [Computer software]. https://cran.r- project.org/web/packages/emmeans/index.html Linné, C. von, Linné, C. von, Smith, J. E., Baillie, M., & London, R. C. of P. of. (1821). A selection of the correspondence of Linnaeus and other naturalists, from the original manuscripts: Vol. v.1 (pp. 1–640). Longman, Hurst, Rees, Orme, and Brown. https://doi.org/10.5962/bhl.title.141275 Liu, X., Zhai, R., Feng, W., Zhang, S., Wang, Z., Qiu, Z., Zhang, J., Ma, F., & Xu, L. (2014). Proteomic analysis of ‘Zaosu’ pear (Pyrus bretschneideri Rehd.) and its early-maturing bud sport. Plant Science, 224, 120–135. https://doi.org/10.1016/j.plantsci.2014.04.012 92 Liu, Y., Gao, X., Tong, L., Liu, M., Zhou, X., Tahir, M. M., Xing, L., Ma, J., An, N., Zhao, C., Yao, J.-L., & Zhang, D. (2022). Multi-omics analyses reveal MdMYB10 hypermethylation being responsible for a bud sport of apple fruit color. Horticulture Research, 9, uhac179. https://doi.org/10.1093/hr/uhac179 Liu, Y., Liu, Q., Xiong, J., & Deng, X. (2007). Difference of a citrus late-ripening mutant (Citrus sinensis) from its parental line in sugar and acid metabolism at the fruit ripening stage. Science in China Series C: Life Sciences, 50(4), 511–517. https://doi.org/10.1007/s11427-007- 0063-8 Migicovsky, Z., Gardner, K. M., Money, D., Sawler, J., Bloom, J. S., Moffett, P., Chao, C. T., Schwaninger, H., Fazio, G., Zhong, G.-Y., & Myles, S. (2016). Genome to Phenome Mapping in Apple Using Historical Data. The Plant Genome, 9(2), plantgenome2015.11.0113. https://doi.org/10.3835/plantgenome2015.11.0113 Morimoto, T., Hiramatsu, Y., & Banno, K. (2013). A Major QTL Controlling Earliness of Fruit Maturity Linked to the Red leaf/Red flesh Trait in Apple cv. ‘Maypole.’ Journal of the Japanese Society for Horticultural Science, 82(2), 97–105. https://doi.org/10.2503/jjshs1.82.97 MSUE. (2014). Growth Stages. Apples. https://www.canr.msu.edu/apples/horticulture/growth- stages Palmer, J. W., Giuliani, R., & Adams, H. M. (1997). Effect of crop load on fruiting and leaf photosynthesis of ‘Braeburn’/M.26 apple trees. Tree Physiology, 17(11), 741–746. https://doi.org/10.1093/treephys/17.11.741 Pratt, H. K., & Goeschl, J. D. (1969). Physiological Roles of Ethylene in Plants. Annual Review of Plant Physiology, 20(1), 541–584. https://doi.org/10.1146/annurev.pp.20.060169.002545 Schneider, G. W., & Childers, N. F. (1941). Influence of Soil Moisture on Photosynthesis, Respiration, and Transpiration of Apple Leaves1. Plant Physiology, 16(3), 565–583. https://doi.org/10.1104/pp.16.3.565 SHAMEL, A. D., & POMEROY, C. S. (1936). BUD MUTATIONS IN HORTICULTURAL 487–494. CROPS. https://doi.org/10.1093/oxfordjournals.jhered.a104171 Heredity, Journal 27(12), of Shane, B., & Lavely, E. (2023). 2023 Predicted Apple Harvest Dates – Grand Rapids Area. Michigan State University Extension. https://www.canr.msu.edu/resources/predicted-apple- harvest-dates-grand-rapids-area Shane, B., Lavely, E., Plotkowski, D., Rothwell, N., Lauwers, E., & Brown, L. (2023, August 9). Michigan apple harvest prediction dates for 2023. Apples. https://www.canr.msu.edu/news/michigan-apple-harvest-prediction-dates-2023 Singh, V., Weksler, A., & Friedman, H. (2017). Different Preclimacteric Events in Apple Cultivars with Modified Ripening Physiology. Frontiers in Plant Science, 8. https://www.frontiersin.org/articles/10.3389/fpls.2017.01502 93 Song, J., Deng, W., Beaudry, R. M., & Armstrong, P. R. (1997). Changes in Chlorophyll Fluorescence of Apple Fruit during Maturation, Ripening, and Senescence. HortScience, 32(5), 891–896. https://doi.org/10.21273/HORTSCI.32.5.891 Sun, H., Abeli, P., Campoy, J. A., Rütjes, T., Krause, K., Jiao, W.-B., Korff, M. von, Beaudry, R., & Schneeberger, K. (2023). The identification and analysis of meristematic mutations within the apple tree that developed the RubyMac sport mutation (p. 2023.01.10.523380). bioRxiv. https://doi.org/10.1101/2023.01.10.523380 Sun, X., Jiao, C., Schwaninger, H., Chao, C. T., Ma, Y., Duan, N., Khan, A., Ban, S., Xu, K., Cheng, L., Zhong, G.-Y., & Fei, Z. (2020). Phased diploid genome assemblies and pan- genomes provide insights into the genetic history of apple domestication. Nature Genetics, 52(12), Article 12. https://doi.org/10.1038/s41588-020-00723-9 Sun, X., Wang, P., Jia, X., Huo, L., Che, R., & Ma, F. (2018). Improvement of drought tolerance by overexpressing MdATG18a is mediated by modified antioxidant system and activated autophagy in transgenic apple. Plant Biotechnology Journal, 16(2), 545–557. https://doi.org/10.1111/pbi.12794 Tartachnyk, I. I., & Blanke, M. M. (2004). Effect of delayed fruit harvest on photosynthesis, transpiration and nutrient remobilization of apple leaves. New Phytologist, 164(3), 441–450. https://doi.org/10.1111/j.1469-8137.2004.01197.x Telias, A., Lin-Wang, K., Stevenson, D. E., Cooney, J. M., Hellens, R. P., Allan, A. C., Hoover, E. E., & Bradeen, J. M. (2011). Apple skin patterning is associated with differential expression of MYB10. BMC Plant Biology, 11(1), 93. https://doi.org/10.1186/1471-2229-11-93 Thammawong, M., & Arakawa, O. (2007). Starch Degradation of Detached Apple Fruit in Relation to Ripening and Ethylene. Journal of the Japanese Society for Horticultural Science, 76(4), 345–350. https://doi.org/10.2503/jjshs.76.345 Thomas, C., Tholl, S., Moes, D., Dieterle, M., Papuga, J., Moreau, F., & Steinmetz, A. (2009). Actin bundling in plants. Cell Motility, 66(11), 940–957. https://doi.org/10.1002/cm.20389 Tian, Y., Thrimawithana, A., Ding, T., Guo, J., Gleave, A., Chagné, D., Ampomah-Dwamena, C., Ireland, H. S., Schaffer, R. J., Luo, Z., Wang, M., An, X., Wang, D., Gao, Y., Wang, K., Zhang, H., Zhang, R., Zhou, Z., Yan, Z., … Yao, J.-L. (2022). Transposon insertions regulate genome-wide allele-specific expression and underpin flower colour variations in apple (Malus spp.). Plant Biotechnology Journal, 20(7), 1285–1297. https://doi.org/10.1111/pbi.13806 Tong, C., Krueger, D., Vickers, Z., Bedford, D., Luby, J., El-Shiekh, A., Shackel, K., & Ahmadi, H. (1999). Comparison of Softening-related Changes during Storage of `Honeycrisp’ Apple, Its Parents, and `Delicious’. Journal of the American Society for Horticultural Science, 124(4), 407–415. https://doi.org/10.21273/JASHS.124.4.407 Trimble, S. (2020, October 14). How to Analyze Photosynthesis in Plants: Methods and Tools. CID Bio-Science. https://cid-inc.com/blog/how-to-analyze-photosynthesis-in-plants-methods- 94 and-tools/ Tukey, H. B., & Young, J. O. (1942). Gross Morphology and Histology of Developing Fruit of the Apple. Botanical Gazette. https://doi.org/10.1086/335103 USDA ERS - Food Availability and Consumption. (2023). https://www.ers.usda.gov/data- products/ag-and-food-statistics-charting-the-essentials/food-availability-and-consumption/ Wang, A., Tan, D., Tatsuki, M., Kasai, A., Li, T., Saito, H., & Harada, T. (2009). Molecular mechanism of distinct ripening profiles in ‘Fuji’ apple fruit and its early maturing sports. Postharvest Biology and Technology, 52(1), 38–43. https://doi.org/10.1016/j.postharvbio.2008.09.001 Wang, Z., Li, G., Sun, H., Ma, L., Guo, Y., Zhao, Z., Gao, H., & Mei, L. (2018). Effects of drought stress on photosynthesis and photosynthetic electron transport chain in young apple tree leaves. Biology Open, 7(11), bio035279. https://doi.org/10.1242/bio.035279 Wei, L., Cao, Y., Cheng, J., Xiang, J., Shen, B., & Wu, J. (2020). Comparative transcriptome analyses of a table grape ‘Summer Black’ and its early-ripening mutant ‘Tiangong Moyu’ identify candidate genes potentially involved in berry development and ripening. Journal of Plant Interactions, 15(1), 213–222. https://doi.org/10.1080/17429145.2020.1760367 Wu, Y., Fu, T., Wang, Z., Jiao, C., Yang, Z., Ali, B., & Zhou, W. (2015). Differential gene expression analysis of early-ripening mutants of grape (Vitis vinifera L.). Scientia Horticulturae, 194, 7–17. https://doi.org/10.1016/j.scienta.2015.07.022 Wünsche, J., & Lakso, A. (2000). Apple tree physiology—Implications for orchard and tree management. https://www.semanticscholar.org/paper/Apple-tree-physiology-implications-for- orchard-and-W%C3%BCnsche-Lakso/f72a3de510269c5f3b726bf2c881a2f5b02f03e8 Zotta, L. (2015). 200 Year and Growing: The Story of Stark Bro’s Nurseries & Orchards Co. (First). Stark Bro’s Nurseries & Orchards Co. 95 APPENDIX Table 1A: Table containing variables fitting ‘Kidd’s D-8’ individual fruits with Weibull equation “y=a+b(1-exp(-((x+d*ln(2)^(1/e)/d)^e))” where x=GDH. FruitID# Eq'n variables R8P5T33 R8P5T36 R8P5T40 R4P7T52 R4P7T54 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 a b c d e -14.7786 -20.2625 243.2489 269.3274 29192.05 27387.42 200089.7 276027.1 18.14616 25.17357 r^2 0.99949 0.999473 n/a n/a n/a n/a n/a n/a -16.0925 -9.81037 278.4696 235.3808 28819.61 30126.13 113178.7 158352.7 9.833075 15.45435 0.999408 0.999462 a b c d e -5.54555 -4.20008 -12.7617 -19.738 -18.1269 183.761 237.5556 246.1911 262.4043 290.3277 30501.12 30039.02 29896.13 27380.27 29501.82 92120.54 40640.32 84197.78 4002470 12166900 8.942774 2.82606 6.800462 371.6084 1149.293 r^2 0.999343 0.999666 0.999665 0.999209 0.999039 a b c d e r^2 a b c d e n/a n/a n/a n/a n/a n/a -23.0221 -20.274 -13.7894 -16.3933 333.8096 297.2722 212.7947 271.6507 27173.04 27726.81 28709.52 29037.17 124123.2 14450900 3897740 468555 11.0133 1430.081 369.2073 43.69957 0.999423 0.999282 0.999291 0.999399 -9.15377 -9.11297 -16.3655 -14.3757 -14.4708 219.2494 241.0538 244.309 234.4832 221.3806 29562.83 27371.7 28529.96 28331.84 28811.46 143134.7 64766.32 358775.2 345939.7 157295.3 14.31445 6.238303 32.94056 32.69629 13.50557 r^2 0.999132 0.999079 0.999419 0.99962 0.999325 a b c d -4.242 -16.586 -16.6445 -12.4736 -13.8515 172.8713 265.9863 240.6024 230.035 235.5481 30843.54 28007.02 27500.45 27666.63 28104.12 82070.28 216000.9 1187640 84235.23 221189.1 96 Table 1A (cont’d) 5 5 e r^2 8.031553 20.27853 116.1411 7.104348 21.35506 0.999211 0.99949 0.999425 0.999072 0.999515 97 Table 2A: Table containing variables fitting ‘Autumn Gala’ individual fruits with Weibull equation “y=a+b(1-exp(-((x+d*ln(2)^(1/e)/d)^e))” where x=GDH. FruitID# Eq'n variables R9P10T77 R9P10T80 R6P6T42 R6P6T44 R6P6T48 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 a b c d e -12.8614 -14.1475 -15.0931 -15.1211 -9.1989 232.6125 263.115 239.0808 228.8246 189.6441 29143.22 30117.17 29290.24 29622.25 29926.87 3.06E+14 279278.5 606937.7 2760980 169138.1 3E+10 25.19976 57.45338 250.2667 15.50311 r^2 0.999184 0.999312 0.999502 0.999388 0.999314 a b c d e -7.78923 -12.0583 -12.8888 -11.0496 -11.8443 261.8401 250.0777 226.4592 239.5882 273.217 32633.49 31215.53 30443.92 30497.87 31142.94 77376.89 166419 321037.7 114836.3 114929 6.190056 14.56693 29.22473 9.832361 10.0197 r^2 0.999572 0.999383 0.999494 0.999388 0.999615 a b c d e -17.5212 -12.6148 -10.5394 -8.58099 -12.163 239.494 221.3435 216.9841 263.4307 260.5754 28512.28 29331.4 30193.9 31942.03 30774.21 4234720 452144.1 102369.7 70251.74 132533.5 386.1333 41.87417 8.928548 5.353924 11.49471 r^2 0.999083 0.999318 0.999447 0.999093 0.999342 a b c d e r^2 a b c d e n/a n/a n/a n/a n/a n/a -8.03232 -12.5872 -12.0979 -9.36054 185.9505 228.0075 227.573 181.8399 29809.46 29853.82 29420.52 29922.99 94156.63 266305.1 357934 1462850 8.006715 24.19293 34.3791 145.3677 0.999066 0.99955 0.999479 0.99927 -11.6483 -9.66838 -2.70397 -15.4767 -10.4614 210.3919 265.0208 196.6732 260.1989 241.3664 29106.24 30742.31 30272.79 30786.03 29346.59 373556.4 89796.92 49294.65 660264.4 4640250 36.26616 7.660358 4.303254 58.41564 493.9297 r^2 0.999373 0.999452 0.99889 0.999138 0.999282 98 Table 3A: Table containing variables fitting ‘September Wonder Fuji’ individual fruits with Weibull equation “y=a+b(1-exp(-((x+d*ln(2)^(1/e)/d)^e))” where x=GDH. Fruit ID variables R9P3T19 R2P10T74 R4P4T27 R4P4T28 R4P4T31 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 a b c d e -8.3302 -37.613 -11.0884 -9.95104 -10.3801 308.3109 289.8988 356.6878 255.6673 411.6456 29214.52 26815.58 34335.18 31439.69 31439.76 71157.84 1189000 58501.54 86628.8 59028.07 6.970579 81.02593 3.691376 7.422976 4.886517 r^2 0.998677 0.99798 0.999474 0.999678 0.999503 a b c d e -22.6952 -35.8326 -10.8807 -13.1166 -13.6269 374.6911 416.7749 272.6775 354.4169 457.9507 28141.88 28932.44 30873.87 31561.11 32134.48 268466.2 157212.3 67050.02 57254.07 61755.47 26.26278 11.89085 5.181052 4.028458 4.847512 r^2 0.999744 0.999233 0.999547 0.999659 0.999251 a b c d e -6.51645 -11.1426 -9.96459 -5.86727 -7.5745 339.6552 350.2616 390.4459 332.1468 497.7085 29093.4 30252 31686.65 32505.12 34607.71 51487.68 59110.07 52855.43 48456.22 48412.21 4.751867 4.802036 4.102293 3.611961 3.30096 r^2 0.99755 0.999769 0.999683 0.999493 0.999684 a b c d e -0.03457 -9.76093 -6.60575 -10.9771 -18.6124 526.6187 333.4435 347.3878 368.839 492.5191 33697.98 30942.34 33579.3 30957.96 29579.18 37642.41 53821.54 47431.99 78415.41 55954.07 2.405745 3.938487 3.050267 7.053881 4.424552 r^2 0.999174 0.999763 0.999726 0.999346 0.99973 a b c d e -13.3363 -31.7364 -18.8093 -10.8147 -12.3349 284.09 393.493 362.8681 430.752 400.9978 29342.37 28532.51 29670.44 32561.48 31164.36 120356.1 140480.3 96315.37 57721.01 66826.44 11.46549 11.43237 8.142745 4.406268 5.683702 r^2 0.99936 0.999338 0.999563 0.999593 0.999651 99 Table 4A: Table containing variables fitting ‘Aztec Fuji’ individual fruits with Weibull equation “y=a+b(1-exp(-((x+d*ln(2)^(1/e)/d)^e))” where x=GDH. Fruit ID# variables R6P5T33 R6P5T34 R3P1T2 R3P1T3 R3P1T5 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 a b c d e 2.933954 -9.11068 -2.75461 -11.0014 -4.06264 355.2165 294.9136 253.1499 417.6879 400.546 39444.97 38386.95 39148.94 38145.64 39519.12 41154.23 79750.99 54424.13 73610.79 51074.45 1.767552 5.043896 3.157899 4.886753 2.725295 r^2 0.996218 0.99943 0.998435 0.999748 0.999488 a b c d e -4.73901 -20.8374 -0.76954 -8.24091 -4.31885 307.5253 417.5925 480.5034 342.5184 534.5191 38523.05 37710.07 45281.81 37399.1 40523.84 54806.45 117528.1 53094.32 66307.71 51145.34 3.071899 7.533184 2.12136 4.160884 2.491646 r^2 0.999741 0.99952 0.998982 0.99963 0.999386 a b c d e -7.10844 -1.7473 -5.1872 -3.90094 -1.13526 410.8042 583.5026 304.0852 295.6473 456.1313 37142.33 40981.3 36823.43 38570.9 38381.94 53461.62 48425.92 59782.02 55110.88 45196.64 3.049857 2.329276 3.812866 3.086562 2.487479 r^2 0.999158 0.999402 0.999223 0.999058 0.999188 a b c d e -8.71883 -8.89516 -4.80469 -2.14058 -1.46562 388.629 354.6148 388.2507 498.2038 393.4741 36880.17 38994.03 37773.68 39497.02 37715.84 57114.61 63423.12 53450.56 48673.18 46581.25 3.457651 3.578128 3.203993 2.555593 2.693918 r^2 0.999493 0.999589 0.999143 0.99939 0.998497 a b c d e -11.4999 -8.99066 -9.81694 -3.58516 -7.67437 484.6034 329.3124 407.1651 357.0502 363.879 38111.47 37173.17 39688.63 41918.71 37612.85 59840.24 60031.99 60660.4 56512.83 57329.49 3.383604 3.459768 3.034833 2.84947 3.317622 r^2 0.999556 0.999422 0.998856 0.999038 0.999478 100 Table 5A: Table containing variables fitting ‘Maslin’ individual fruits with Weibull equation “y=a+b(1-exp(-((x+d*ln(2)^(1/e)/d)^e))” where x=GDH. Fruit ID Fruit Variables R6P8T61 R6P8T62 R6P8T63 R3P9T69 R3P9T70 a b c d e 0.06866 -1.16861 -4.50129 -0.99457 -1.48408 1323.415 448.2086 253.3919 417.0138 309.6818 142904.5 55133.52 38629.48 39680.29 38875.98 177124 65002.54 54400.88 47081.22 47776.58 1.474805 1.915038 2.68527 2.344374 2.51705 r^2 0.999596 0.999693 0.999277 0.999464 0.999521 a b c d e r^2 a b c d e r^2 a b c d e -0.33123 513.9362 51550.89 59978.71 1.692566 0.999664 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a 0.790002 -2.4533 0.350288 320.0965 438.2541 514.0266 44479.65 40611.47 56737.21 52309.13 49554.04 65820.79 2.001848 2.355074 1.737849 0.999548 0.99957 0.999152 0.380962 -2.61855 -5.98574 -8.50168 578.7336 314.8147 277.1088 309.896 65338.22 42439.71 40119.53 39721.1 76448.87 52748.1 64078.91 62349.57 1.646007 2.286415 3.616752 3.245363 0.999501 0.999235 0.999341 0.999017 -8.08906 0.527276 311.5888 394.991 47507.99 45238.44 69891.68 51316.31 2.621621 1.86293 n/a n/a n/a n/a n/a n/a -6.15839 -3.53778 587.928 302.807 63060.04 37674.11 81015.87 50734.34 2.116428 2.831615 0.999598 0.999577 r^2 0.999594 0.999269 a b c d e -2.77432 0.73366 -2.75421 -7.21217 -8.80547 454.9381 443.3501 306.202 302.0484 306.7224 67291.35 47091.75 40625.5 37494.76 37050.62 83847.33 52378.81 51217.11 58371.82 61136.91 1.839799 1.825036 2.382964 3.333166 3.464483 r^2 0.999278 0.999535 0.999759 0.999514 0.999573 101 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 Table 6A: Table containing variables fitting ‘Cripps Pink’ individual fruits with Weibull equation “y=a+b(1-exp(-((x+d*ln(2)^(1/e)/d)^e))” where x=GDH. Fruit ID Fruit Variables R5P7T52 R5P7 T54 R6P10T77 R6P10T78 R6P10T79 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 4 4 4 4 4 4 5 5 5 5 5 5 a b c d e -5.62069 -7.74034 -8.04981 -8.27267 -19.3067 304.4466 373.0066 290.805 245.2403 228.2765 39273.99 41497.07 45606.86 43694.44 40484.64 59154.74 60237.17 65557.75 70401.57 204434.6 3.141581 2.977606 2.412422 2.927474 10.056 r^2 0.99967 0.999734 0.999596 0.99959 0.999579 a b c d e -6.63533 -12.486 -8.27004 349.2691 304.038 314.2441 44976.79 40612.97 48797.97 63589.6 69587.51 72925.42 2.625894 3.133108 2.614779 r^2 0.99988 0.999691 0.999579 n/a n/a n/a n/a n/a n/a -21.3761 228.7415 36069.79 441884 25.51886 0.999617 a b c d e r^2 a b c d e r^2 a b c d e -1.36917 416.1578 47083.6 57474.72 2.129991 0.999451 -3.55025 301.5525 40550.15 52798.24 2.512841 0.99965 n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a -1.5162 -6.58313 -9.28882 271.8481 292.1664 187.23 51465.99 44745.92 38319.77 62871.29 64170 88361.75 1.834832 2.646137 4.920321 0.999396 0.99948 0.999283 -8.90982 -24.333 -3.16585 214.0246 264.8133 332.5045 40843.82 39448.53 50694.27 75268.05 205574.4 63286.11 3.581775 9.980935 1.977023 0.999595 0.999536 0.999135 -23.6679 -0.00865 -12.6905 -14.0427 -24.7252 318.3164 349.8363 226.6002 215.1272 241.9423 43014.81 49971.74 40560.02 40250.17 37859.51 123743.5 57025.96 102154.4 108264.7 329208.6 5.383155 1.946425 5.079653 5.217302 16.8965 r^2 0.999523 0.999703 0.999505 0.99955 0.999676 102 Figure 1A: Carbon assimilation shown for each cultivar. Each point represents the average of 2 leaves per tree, five trees in total. Data was analyzed via ANOVA in R software (R Core Team, 2017). Temperature data collected from Clarksville weather station. Results of carbon assimilation comparisons between early and late cultivars are mixed. 103 Figure 2A: Carbon assimilation shown for each cultivar, capturing differences in carbon assimilation throughout the season according to developmental stage and maturity (earlier or later harvesting cultivar). Each different color represents a different date when the data was collected: red=June 16, brown=July 27, green=18 August, and pre/postharvest dates are unique to cultivar and shown in Fig. 1A above within the appendix. Each data point represents an average of 2 leafs per 5 trees. Each complete area under the diurnal curve represents total carbon assimilated for that particular date/developmental timepoint. 104