GENETIC CHARACTERIZATION OF RESISTANCE TO PHYTOPHTHORA CAPSICI AND 
MORPHOLOGICAL DIVERSITY IN CUCUMBER 

By 

Ying-Chen Lin 

A DISSERTATION 

Submitted to 
Michigan State University 
in partial fulfillment of the requirements   
for the degree of 

Plant Breeding, Genetics and Biotechnology – Horticulture - Doctor of Philosophy 

2023 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ABSTRACT 

Cucumber (Cucumis sativus L.) is an economically important vegetable crop that is 

cultivated around the world. Pickling cucumber production in the United States is greatly 

impacted by Phytophthora fruit rot, caused by a soil-born oomycete pathogen, Phytophthora 

capsici. With no resistant variety available, genetic resources are needed to develop resistant 

varieties. The goal of this work was to identify quantitative trait loci (QTL) associated with 

resistance to Phytophthora fruit rot using multiple genomic approaches and populations. Young 

fruit resistance, which is observed during the state of rapid fruit growth prior to harvest, is a 

quantitative trait for which multiple QTL were identified. The largest effect QTL, qPFR5.1, 

located on chromosome 5 was fine mapped to a 1-Mb region. Genome-wide association studies 

(GWAS) and extreme-phenotype genome-wide association study (XP-GWAS) for young fruit 

resistance also were performed on a resequenced cucumber core collection representing > 96% of 

the genetic diversity of the USDA cucumber germplasm. Several SNPs overlapped with the QTL 

identified from QTL-seq analysis on biparental populations. In addition, novel SNPs associated 

with the resistance were identified from the germplasm. The resistant alleles were found mostly in 

accessions from India and South Asia, the center of origin for cucumber. The results from this 

work can be applied to future disease resistance studies and marker-assisted selection in breeding 

programs. When introgressing disease resistant traits, the resulting varieties also need to meet 

market standards, which are defined by morphological characteristics. Disease resistance and 

other valuable traits are often found in landraces or wild accessions that have fruit quality traits 

that are undesirable in commercial markets. To provide morphological and genetic information 

for the diversity of fruit traits in the cucumber core collection, external and internal fruit quality 

traits including fruit shape, skin color, netting, spine density, curvature, seed cavity size, flesh 

 
 
thickness, hollowness, and flesh color, were measured in 2019-2022. The photographic and 

phenotypic data are deposited and publicly available on the Cucurbit Genome Database 

(GuGenDB, http://cucurbitgenomics.org/v2/). GWAS analyses were performed on all traits 

collected. The findings from the association analyses supported many previously identified QTL 

and candidate genes, suggested additional candidate genes, and provide genetic framework for 

morphological diversity of the core collection. 

 
 
 
 
ACKNOWLEDGEMENTS 

The long journey has come to an end. Looking back now, I was very lucky and fortunate 

to be in this place and met many wonderful people. I have myriad joyful memories here in MSU, 

and I sure will miss it very much.  

This research cannot be completed without the help and support from many people. First 

of all, I have to thank my PhD advisor, Dr. Rebecca Grumet – you cannot ask for a better mentor 

for a grad student. I am so grateful for your support and guidance during my PhD study, whether 

during hours of brainstorming research ideas in the kitchenette, or in the field and greenhouse for 

countless harvests starting at 7 am.  

I also must thank my committee members, Dr. Robert VanBuren, Dr. Karen Cichy and Dr. 

Ray Hammerschmidt, for their advice and encouragements. Big thanks to all the Grumet lab 

members: Sue Hammar, Stephanie Rett-Cadman, and Ben Mansfeld. I am so glad to have all of 

you in this lab, you are the most bright and delightful people I met, we always had good times 

even when things didn’t go well (This is fine!). Also, definitely can’t leave out all the 

undergraduate students who helped with all the sweaty field and greenhouse work.  

I would also like to joint lab group for the opportunities to learn and share interesting 

research from everyone and give me valuable feedback on my work. And thank you to all the 

fellow graduate students for the support and friendship we share. Thank you to administrative 

staff of the Horticulture Department, being always cordial and help with all the questions I 

brought to them. I also want to thank all the staff in the Plant Science Greenhouses and 

Horticulture Teaching and Research Center for assisting with even the smallest requests I had and 

make sure my plants grew happily.  

Finally, I want to thank my parents. Being 12-hour time difference apart is hard, but 

 iv 

 
always give me their unwavering support and belief, and always let me know how they’re very 

proud of me. This thesis is dedicated to them. 

 v 

 
 
 
 
TABLE OF CONTENTS 

CHAPTER 1  Literature Review ................................................................................................. 1 
Cucumber production and domestication history ........................................................................... 2 
Genomic tools for cucumber .......................................................................................................... 5 
Resistance to Phytophthora fruit rot ............................................................................................... 7 
Morphological diversity in the cucumber core collection ............................................................ 13 
WORKS CITED ........................................................................................................................... 24 

CHAPTER 2   Identification of QTL Associated with Resistance to Phytophthora Fruit Rot 
in Cucumber (Cucumis sativus L.) ............................................................................................. 37 
Abstract ......................................................................................................................................... 38 
Introduction ................................................................................................................................... 38 
Materials and Methods .................................................................................................................. 41 
Results ........................................................................................................................................... 50 
Discussion ..................................................................................................................................... 66 
Conclusions ................................................................................................................................... 73 
WORKS CITED ........................................................................................................................... 75 
APPENDIX A: Supplementary Materials .................................................................................... 81 

CHAPTER 3  Phenotypic and Genetic Characterization of Morphological Diversity in the 
Cucumber Core Collection ......................................................................................................... 85 
Introduction ................................................................................................................................... 86 
Materials and Methods .................................................................................................................. 88 
Results ........................................................................................................................................... 92 
Discussion ................................................................................................................................... 105 
Conclusion .................................................................................................................................. 111 
WORKS CITED ......................................................................................................................... 113 
APPENDIX B: Supplementary Materials .................................................................................. 120 

CHAPTER 4  Conclusion and Future Directions ................................................................... 121 

 vi 

 
 
 
 
 
 
 
CHAPTER 1 

Literature Review

1 

 
 
Cucumber production and domestication history 

Domestication history 

Cucumber has been widely cultivated around the world following domestication ~3000 

years ago (Sebastian et al., 2010; Renner and Schaefer, 2016). The origin of cucumber is thought 

to be in India, China, Burma, Thailand, and the primary and secondary centers of diversity are in 

India and Southeast Asia, respectively ((McCreight et al., 2013; Naegele and Wehner, 2016). 

Following initial domestication, cucumber was introduced to China ~2,000 years ago, and was 

further introduced to East Asia (e.g., Japan) around ~100 CE (Lv et al., 2012). It was also 

dispersed westward to Europe and Africa through two routes: Persia into eastern and northern 

Europe by land in 6th or 7th century, and from Persia to western and southern Europe by sea in 

10th century (Staub et al., 2008; Paris et al., 2012). The plant was later brought to the Americas 

by Columbus in 1490s.  

Cucumber taxonomy 

The cucumber species can be categorized into four botanical varieties: wild cucumber (C. 

sativus var. hardwickii), Sikkim cucumber (C. sativus var. sikkimensis Hook. F), semi-wild 

Xishuangbanna cucumber (XIS, C. sativus var. xishuangbannanesis Qi et Yuan), and cultivated 

cucumber (C. sativus var. sativus).  

The wild cucumber, C. sativus var. hardwickii (Royle) Alef (HARD) is distributed widely 

in north-western Himalayan mountains. The plant is used as a laxative by native people (Deakin 

et al., 1971). The plant has strong branching habit, and produces a large number of small oval-

shaped fruits (4-5 cm), which taste bitter and drop easily when mature (Deakin et al., 1971; Weng, 

2021). The tiny seeds have strong dormancy, which require treatments such as removing seed 

coats to increase germination rates (Weston et al., 1992).  

2 

 
XIS cucumber is widely distributed in Xishuangbanna of southwest China and 

surrounding regions (Thailand, Laos, and Myanmar) and is mainly consumed by locals (Qi et al., 

1983; Yang et al., 1991). The non-bitter fruit of XIS cucumber are large, mainly have five carpels, 

and produce seeds with strong dormancy (Qi et al., 1983; Renner, 2017; Weng, 2021). The 

distinct feature of XIS cucumber is orange flesh that is rich in β-carotene (Navazio and Simon, 

2001; Bo et al., 2012). Genetic analysis showed that after domestication, XIS cucumber 

underwent diversification selection for unique traits (Bo et al., 2015). The inheritance and genetic 

mechanism of β-carotene accumulation was well studied and has been bred to increase nutrition 

level (Cuevas et al., 2010; Qi et al., 2013).  

Sikkim cucumber is found in the Sikkim region of India and Nepal (Hooker, 1876). The 

plant produces large and bulky fruits (up to 38 cm in length and 12 cm in diameter) in brown 

color with netting, as well as a large centric hollow when mature. In addition to these distinct 

features, the Sikkim cucumber also have many lateral branches, later flowering and seeds with 

different degrees of dormancy (Wang et al., 2021b). Many accessions with similar morphological 

features also called Sikkim cucumber, also have been collected from Nepal, Bhutan and northern 

India (Sharma and Hore, 1996; Wang et al., 2021b; Yashiro et al., 2017). 

Through natural and artificial selection, cucumber shows great differences compared to 

the wild progenitor. For example, cultivated cucumber has fewer lateral branches and bears fewer 

but larger fruits with fewer spines, thicker flesh, smaller seed cavity size, no intercentrum hollow, 

and loss of bitterness (Harlan, 1992; Walters et al., 1996; Abbo et al., 2014; Weng, 2021). The 

plants also have larger seeds without dormancy, which can facilitate the cultivation process. 

During the process of selection, cucumber underwent narrow bottlenecks which led to loss of 

diversity (Qi et al., 2013). Despite the overall trend of selection, the morphology of cucumber 

3 

 
varies depend on local environments and consumer preferences. As result, cultivated cucumber 

has been bred for specific traits and markets (Weng, 2021; Grumet et al., 2023). 

Market classes of cucumbers 

Consumer preferences and handling requirements for commercial production has 

deepened the distinct phenotypes between different market classes. Based on the use, whether 

consumed fresh or processed, and the different target markets (Figure 1.1), the types of cucumber 

can be categorized into several major market classes. In East Asia, long and slender cucumber are 

preferred and mostly consumed fresh. They can be further classified into North and South China 

types. The North China type, also called “Chinese Long”, typically has long and  tapered fruit 

covered with dense white spines, while the South China type has cylindrical fruit with green and 

white stripes and fewer black spines (Weng, 2021). North American and Europe fresh market 

cucumbers (slicing cucumbers) have smooth skin and intermediate fruit length (20-30 cm), while 

the processing cucumbers (pickling cucumbers) tend to have thicker and warty fruit surface, and 

are generally short in length (5-15 cm). Another major market type is parthenocarpic cucumber, 

typically produced in greenhouses, which includes two classes: long-fruited, about 30-40 cm long, 

and Beit Alpha (Mediterranean) cucumber, which has shorter fruit (12-15 cm) (Weng, 2021; 

Grumet et al., 2023).  

Cucumber production in the U.S. 

In the United States, the cucumber production primarily serves two different markets: 

slicing (fresh) and pickling (processing). Michigan is one of the leading pickling cucumber 

producing states with more than 171,000 tones harvested in 2015. In commercial pickling 

cucumber production in Michigan, cucumber plants are grown on the ground in high density to 

facilitate machine harvest (Michigan State University Extension, 

4 

 
http://msue.anr.msu.edu/news/pickling_cucumber_planting_density_affects_yield_and_dollar_val

ue). Unlike hand-picked cucumbers, which allows multiple harvests to increase yield but requires 

higher labor cost and more field space, machine-harvested cucumbers are usually grown at higher 

density with 24-28 inches between row and 3-5 inches between plants within the row. Machine 

harvest is a once-over, large-scale, and destructive operation that usually harvests 1-1.5 fruit per 

plant on average. Pickling cucumbers are harvested when they are less than 5-1/2 inches in length 

or 1-7/8 inches in diameter (USDA Pickling Cucumbers Grades and Standards, 

https://www.ams.usda.gov/grades-standards/pickling-cucumbers-grades-and-standards). There are 

several diseases that can significantly impact the production of cucumber, such as bacterial wilt, 

powdery mildew, downy mildew, angular leaf spot, and Phytophthora fruit rot. Among these, the 

most severe diseases that impact cucumber production in Midwestern US , are downy mildew and 

Phytophthora fruit rot (Hausbeck and Lamour, 2004; Savory et al., 2011). 

Genomic tools for cucumber 

Reference genomes  

Cucumber was the first sequenced cucurbit crop (Huang et al., 2009). Several 

characteristics make the plant suitable for facilitating genetic and genomic studies: it is diploid, 

has a low chromosome number (2n = 2x = 14), a relatively small genome size (~400 Mbp), low 

percentage of repetitive sequences, and a short life cycle. Currently, three draft genomes have 

been published from the cucumber research community: the North China type ‘Chinese Long’ 

inbred line 9930 (v3.0) (Li et al., 2019), European inbred line ‘B10’ (v3.0) (Osipowski et al., 

2020), and the North American pickling type inbred line ‘Gy 14’ (v2.1) (Yu et al., 2023). All the 

reference genomes are available on the Cucurbit Genome Database (GuGenDB, 

http://cucurbitgenomics.org/v2/). With draft genomes and new genomic tools available, genetic 

5 

 
linkage maps have been generated from different populations targeting various traits of interests 

such as growth and flowering traits, yield, fruit shape and quality, and resistances to abiotic and 

biotic stresses. The constructed linkage maps, coupled with QTL identification, can ultimately 

boost marker-assisted breeding and map-based gene cloning.  

Figure 1.1. Examples of cucumber market types. Left to right: Western fresh market (slicing); 
Beit Alpha/Mediterranean; parthenocarpic greenhouse; Western pickling; Chinese Long. (Figure 
from Grumet, Lin, et al., 2023). 

The CucCAP project 

The cucumber germplasm resources for breeding and genetic study are maintained in 

several collections worldwide including the USDA Agriculture Research Service National Plant 

Germplasm System (NPGS), Institute of Vegetables and Flowers at the Chinese Academy of 

Agricultural Sciences in China, and the Centre for Genetic Resources in the Netherlands. The 

NPGS stores 1,314 cucumber Plant Introduction (PI) accessions, which mainly are cultivars, 

6 

 
 
 
 
landraces, and varieties. The USDA-SCRI CucCAP project: Leveraging Applied Genomics to 

Increase Disease Resistance in Cucurbit Crops, aims to develop genomic tools for disease 

resistance for the crops in the Cucurbitaceae family (Grumet et al., 2020). To assess genetic 

diversity and relationships among accessions, the NPGS cucumber collection was sequenced 

using genotyping-by-sequencing (GBS) (Wang et al., 2018). Phylogenic analysis of the collection 

showed three distinct clades that reflect cucumber domestication history: India/South Asia, East 

Asia, and the rest of the world (Central/West Asia, Europe, Africa, and the Americas) (Lv et al., 

2012; Qi et al., 2013; Wang et al., 2018). Within the major clades, there are subclades that showed 

regional divergence and morphology differences, for example, separation of accessions from 

eastern and western Türkiye, as well as fresh and pickling market classes in North American 

accessions (Wang et al., 2018; Grumet et al., 2020).  

Based on the GBS sequence of the NPGS cucumber collection, a core collection of 388 

accessions was established (Wang et al., 2018). The core collection captures > 95% of the allelic 

diversity in the NPGS cucumber collection and also includes historical cultivars with important 

agronomic traits and disease resistance (Wang et al., 2018). To reduce heterozygosity and 

heterogeneity within each accession, the accessions were self-pollinated for 2-3 generations. The 

self-pollinated core collection accessions were re-sequenced at 30-40× coverage and used to call 

SNPs as described by Yu et al. (2023). The SNP data is accessible on the Cucurbit Genomics 

Database website (CuGenDB) (http://cucurbitgenomics.org/v2/).  

Resistance to Phytophthora fruit rot 

Phytophthora blight, caused by the soil-borne oomycete pathogen Phytophthora capsici 

(Leon.), is one of the most serious concerns for cucurbit growers in the eastern and Midwestern 

United States. The pathogen can spread to other cucurbit growing areas through irrigation and 

7 

 
rainwater, and once present the oospores can linger in soil for many years. In very severe cases it 

can lead to 100% yield loss. Currently, there is no single effective management strategy to control 

the disease, preventative measures such as planting in clean fields and avoiding contaminated 

water sources are the most effective methods. However, once the field is infested, growers have 

no choice but to switch to other non-host crops. Since P. capsici can infect Cucurbitaceous, 

Fabeceous, and Solanaceous crops, farmers are left with a finite choice of crops.  

Life cycle 

The thick cell wall of P. capsici oospores protects it from desiccation and other extreme 

environmental conditions such as cold temperature. Under moist and warm conditions, oospores 

germinate and grow into hyphae, with lemon-shaped sporangia on the external plant surface 

(Lamour, 2013). Sporangia can release motile biflagellate zoospores in water, zoospores then 

germinate when in contact with the plant surface. When they perceive a suitable host surface, the 

hyphae grow out from zoospores and form appressoria, a specialized infection structure to breach 

the plant cuticle and cell walls. As a hemi-biotrophic pathogen, P. capsici first lives biotrophically 

in plant tissue. After penetrating the plant surface, some hyphae that grow within the plant tissue 

will form haustoria at the tip to acquire nutrients from host plants; often with no visible 

symptoms. In later infection stages, P. capsici kills host cells which results in necrosis and tissue 

collapse, followed by sporulation on the host surface which then can start a new infection cycle 

(Lamour et al., 2012).  

Disease controls and prevention 

In commercial cucumber production, fruits are located in contact with the ground under a 

canopy of leaves; the direct contact to soil increases the probability of P. capsici infection (Ando 

and Grumet, 2006). In addition to soil contact, the other major infection pathway is through 

8 

 
irrigation and rainwater, which help P. capsici to release mobile zoospores to spread within or 

between fields (Granke et al., 2012). In cucumber, P. capsici specifically infects fruits, especially 

young fruits, while leaves and vines remain healthy (Gevens et al., 2006). In addition, the latent 

period before obvious symptoms appear can result in serious economic loss, as infected fruits that 

lack symptoms can be harvested and later infect other fruits during storage. 

There is no single management that can effectively control Phytophthora fruit rot, a 

combination of exclusion, cultural practices and chemical control strategies must be used to 

reduce epidemics of P. capsici. Among all the methods, avoiding the pathogen by planting in the 

fields that have no P. capsici infection history is the most crucial (Babadoost, 2005). Trellises, 

adding plant space, and cover crops are possible approaches to reduce disease level, however, 

these are not practical in large-scale commercial pickling cucumber production (Ando et al., 

2009). The wide host range, genetic variance between isolates, and long-term survival of oospores 

increase the difficulty of disease management. Additionally, resistance to a commonly used 

fungicide, mefenoxam, has been reported in several states from different crops (Lamour and 

Hausbeck, 2000; Parra and Ristaino, 2001; Café-Filho and Ristaino, 2008; Davey et al., 2008; 

Dunn et al., 2010). Longevity of the oospores in soil also results in failure of non-host crop 

rotation although rotation can decelerate the buildup of disease pressure (Hausbeck and Lamour, 

2004). A study conducted in Illinois showed that P. capsici can survive in soil for more than 36 

months and still remain virulent (Babadoost and Pavon, 2013). Apart from the disease 

management strategies mentioned above, host resistance is the most desirable strategy for P. 

capsici management. Growing a resistant variety also can reduce the use of fungicides, which lead 

to lower cost and less potential environmental or health hazards. However, at this time there are 

no genetically resistant cucumber varieties available.  

9 

 
Inheritance and mechanisms and of host resistance to P. capsici   

Efforts have been undertaken to introgress genes for resistance to P. capsici into 

commercial varieties or discover new resistance genes residing in related wild species for 

Solanaceous crops and some cucurbit crops. In watermelon, four accessions from Citrullus 

lanatus var. lanatus showed high resistance to fruit rot and one C. colocynthis and a C. lanatus 

var. citroides also exhibited resistance, but to a lower degree (Kousik et al., 2012). Broad 

resistance to Phytophthora fruit rot observed during post-harvest was identified in five 

watermelon germplasm lines (Kousik et al., 2022), and a breeding line, USVL531-MDR, resistant 

to both powdery mildew and Phytophthora fruit rot was recently released (Kousik et al., 2022, 

2023). Screening of a collection of 308 melon (Cucumis melo) PIs and two commercial cultivars 

for P. capsici crown rot resistance identified three PIs that were highly resistant (Donahoo et al., 

2013). Another screening for P. capsici resistance that was specifically focused on melon 

seedlings, identified two accessions that showed no symptoms in response to five P. capsici 

isolates from different hosts (Pontes et al., 2014). Several studies have identified multiple PIs in 

Cucurbita moschata and Cucurbita pepo L. (squash, pumpkin, and gourd) with crown rot 

resistance (Padley, Les D. et al., 2009; Krasnow et al., 2014; Michael et al., 2019; LaPlant et al., 

2020). The sources identified could be used to expand available resistant germplasm and can be 

integrated to future breeding programs. QTL associated with P. capsici resistance have also been 

identified in multiple cucurbit crops. In melon, a QTL on chromosome 12 was identified, for 

which a candidate gene encoding a wall-associated receptor kinase (WAK) was proposed (Wang 

et al., 2020b). For squash, six QTL in C. pepo (Vogel et al., 2021) were identified using bulked 

segregant analysis (BSA) and four QTL were identified in C. moschata using BSA and QTL 

mapping approaches (Ramos et al., 2020; Michael et al., 2021).  

10 

 
In some crop species, organ-specific resistance is observed. In pepper, roots, stems, 

foliage, and fruit are all susceptible. In tomato, crown rot, leaf spot, and foliar blight are seen 

(Quesada-Ocampo and Hausbeck, 2010; Quesada-Ocampo et al., 2016). In cucumber, P. capsici 

mainly infects fruits, particularly young fruits, while leaves and vines remain largely unaffected 

(Gevens et al., 2006). In watermelon, similar to cucumber, fruit rot is more severe while foliar 

symptom is limited to water soaking, and in summer squash, both leaf and fruit are susceptible 

(Kousik et al., 2017; Quesada-Ocampo et al., 2023). Walker and Bosland, (1999) reported the 

organ specific expression of genes related to P. capsici resistance in pepper. They found that one 

independent dominant gene was essential for root rot resistance and another one for foliar blight 

resistance, with the presence of at least one dominant allele from the third gene. In addition to 

Phytophthora root rot and foliar blight, Sy et al. (2005) found an additional symptom, stem blight 

resistance, was controlled by one dominant gene. They concluded that there were three different 

dominant genes responsible for the resistance on each symptom (root rot, foliar blight and stem 

blight). Subsequent research conducted by Ogundiwin et al. (2005) supported these findings, 

however, instead of a single dominant gene, they identified QTLs from two recombinant inbred 

line (RIL) populations that had different effects on root and/or leaf symptoms. 

P. capsici resistance in cucumber fruit 

Age-related resistance 

From the initial search for genetic resistance to P. capsici in cucumber, a population of 

333 cucumber cultigens was tested (Gevens et al., 2006). However, instead of complete 

resistance, an age-related resistance (ARR) was observed that corresponds with fruit growth 

(Gevens et al., 2006). During the first few days after pollination, about 0-5 days-post-pollination 

(dpp), cucumber ovaries undergo cell division, which is then followed by a period of rapid cell 

11 

 
expansion, usually lasting until 12-16 dpp (Ando et al., 2012). Cucumber fruits are usually 

harvested immature, approximately 8-12 dpp, during mid to late exponential growth. Finally, fruit 

and seeds mature at ~30-35 dpp. Cucumber fruits are susceptible to P. capsici in the early fruit 

development stages, then gradually become resistant. The transition from susceptible to resistant 

begins at ~12 dpp, toward the end of exponential growth (Ando et al., 2009). ARR is also 

observed in other cucurbit crops such as watermelon and various squashes (Ando et al., 2009; 

Krasnow et al., 2017; Alzohairy et al., 2020, 2021). 

Studies of cucumber fruit peels indicated that the fruit surface is important for ARR, and 

methanolic peel extracts of resistant-age peels suggested that biochemical components in the fruit 

peel may contribute to the ARR response (Ando et al., 2015). An untargeted metabolomics 

analysis showed a specific increase of terpenoid glycosides in resistant-age peels and 

transcriptomic analysis of the peel tissues identified enrichment of genes involved in terpene and 

flavonoid synthesis pathways (Mansfeld et al., 2017). Fruit exhibiting ARR also showed the 

ability to respond rapidly to the presence of P. capsici zoospores, as evidenced by a spike in 

defense response genes relative to susceptible-age fruit as early as 2 hours post inoculation and 

observable death of zoospores as early as 4 hour-post-inoculation (Mansfeld et al., 2020).  

Young fruit resistance 

Since ARR does not take effect until the end of exponential fruit growth, which is after the 

time for harvest, cucumber growers will still suffer yield loss due to P. capsici infection, even 

with the cultivar exhibit ARR. Given the need to protect young fruits, Colle et al. (2014) screened 

the full cucumber U.S. cucumber PI collection for young fruit resistance to P. capsici. Of 1,076 

accessions screened, three were identified as potential sources of resistance to young fruit 

resistance. One accession, PI 109483 collected from Türkiye in 1935, showed stable resistance 

12 

 
over several generations of selfing. PI 109483 also has been found to be a potential source of 

resistance to belly rot caused by Rhizoctonia solani (Uchneat and Wehner, 1998). 

Morphologically, PI 109483 has dark green fruit that turns yellow when mature, the fruit shape is 

oblong with smooth skin and minor striping. A breeding line derived from PI 109483 was 

released as MSU 109483-53 (Grumet and Colle, 2017). 

Objectives 

The first objective of this thesis was to identify QTL associated with young fruit resistance 

to P. capsici and develop molecular markers to facilitate future introgression and breeding line 

development. A combination of genomic approaches, including QTL-seq, fine mapping, GWAS 

and XP-GWAS, and population types, including bi-parentally-derived populations and a diversity 

panel (the cucumber core collection) were utilized in this work. 

Morphological diversity in the cucumber core collection 

Fruit appearance is one of most important aspects that determines commercial value of 

cucumber. Key cucumber fruit quality traits include both external and internal features such as 

size and shape, flesh thickness, seed cavity size, and lack of internal hollowness; and skin features 

including color, spines and glossiness. The cucumber core collection exhibits extensive diversity 

in fruit traits including both desirable and undesirable features (Figure 1.2). 

While a great deal of breeding research focuses on resistance to abiotic and biotic stresses 

to cope with existing and emerging challenges, the resulting breeding lines and varieties need to 

meet market requirements for fruit type. Many sources of disease resistance are found in wild or 

landrace accessions, which usually have more feral and undesirable morphology. For instance, a 

Sikkim type cucumber from India (PI 197087) was found to be resistant to downy mildew, 

powdery mildew, anthracnose and angular leaf spot (Wang et al., 2021b). However, the bulky, 

13 

 
netted, and large centric hollow fruit of Sikkim cucumber are highly unfavorable traits. To 

incorporate these resistant traits into commercial lines may require additional cycles of 

backcrossing to develop a morphologically acceptable variety. Selecting appropriate breeding 

materials can possibly achieve breeding goals with shorter breeding generations. Information 

about genetic regulation of the traits can facilitate these efforts. To this date, extensive research 

has been performed on bi-parental populations to identify QTL and candidate genes for various 

fruit quality traits, but few were performed on diversity panels.  

Genetic control of cucumber fruit external traits 

Fruit size and shape 

Cucumber fruit size and shape related QTL have been well studied in numerous 

populations as summarized and reviewed in several papers (e.g., Weng 2015; Che and Zhang, 

2019; Pan et al., 2020; Wang et al., 2020;  Zhang et al., 2021a; Weng 2021;and Grumet et al. 

2023). A set of 21 consensus fruit size and shape related QTL that were recurringly identified in 

different populations and studies was summarized by Weng et al. (2015).  

Fruit shape is defined by the ratio between fruit length and diameter, which is one of the 

crucial attributes for market value. Many fruit shape candidate genes have been identified. The 

candidate gene for the QTL FS1.2, CsSUN, had a similar effect on fruit shape to its homolog gene 

SUN in tomato (Pan et al., 2017). A candidate gene for QTL Fl4.1 encoding a homolog of an 

auxin efflux carrier family protein was downregulated in long-fruited lines (Xing et al., 2023). 

The gene CRABS CLAW (CsCRC) was found to be the candidate gene for QTL FS5.2, which 

resulted in longer fruit shape, potentially mediating auxin and gibberellic acid (GA) signaling 

pathways (Pan et al., 2022). The rare allele CsCRCA associated with short fruit only existed in 

round-fruited semi-wild Xishuangbanna cucumbers, while the alternative allele CsCRCG  can be 

14 

 
found in both wild and cultivated cucumber (Che et al., 2023). 

A cloned gene, CsTRM4 (also named CsTRM5, TONNEAU1 Recruiting Motif5), which is 

a homolog of tomato SlTRM5, was identified for FS2.1 QTL (Wu et al., 2018). Knocking out the 

expression of CsTRM5 (CsaV3_2G013800) repressed cell expansion and resulted in shorter fruit 

length (Xie et al., 2023). Several short fruit mutants (sf1, sf2, sf3, and sf4) were cloned and  

Figure 1.2. Examples of diversity in cucumber ovary and fruit morphology at different 
stages of development as observed in the CucCAP cucumber core collection. (A) Cucumber 
ovaries at anthesis or one day post-anthesis exhibit differences in size, shape, 
presence/absence/number of spines, spine color, warts, and ribs. (B) Variation in cucumber fruit 
development during early exponential growth 5–8 days post pollination (dpp). (C) Mature 
cucumber fruit vary in shape, size, color, surface texture, and netting. (D) Variation in internal 
properties at maturity (flesh thickness, seed cavity size, color, and hollows). (Figure from Grumet, 
Lin et al., 2023).  

15 

 
 
 
 
cucurbit-specific RING-type E3 ligase that regulates fruit elongation by mediating a key enzyme 

characterized from multiple mutagenesis populations. The short fruit 1 (sf1) gene encodes a fourth 

short fruit mutant, sf4, encodes a homolog of an O-linked N-acetylglucosamine (GlcNAc) 

transferase (OGT), influences gene expression involved in cell division in Arabidopsis (Zhang et 

al., 2023). The CsFUL1A allele of a FRUITFULL (FUL)–like MADS-box gene, which can be 

readily found in East Asia accessions with longer fruits, influences auxin accumulation and cell 

division and expansion (Zhao et al., 2019).  

Curvature and Tapering 

To integrate with modern large-scale commercial production, uniform and straight 

cucumber fruits are required to streamline packing and shipping. However, some accessions have 

higher tendency to produce curved or tapered fruit. Fruit curvature can be caused by an 

unbalanced distribution of auxin, where the convex side has higher auxin concentration than the 

concave size (Li et al., 2020). Applying auxin or overexpressing of CsYUC10b, a YUCCA 

involved in auxin biosynthesis, promoted auxin accumulation symmetrically on both sides of the 

fruits, and resulted in straight fruits (Li et al., 2020). A mutant exhibiting a tapered-shaped fruit 

phenotype (mango fruit, mf) that has underdeveloped seed cavity and carpel separation was 

caused by a mutant in a WUSCHEL-related homeobox1 (CsWOX1) gene (Niu et al., 2018).  

Fruit skin color 

Since cucumber fruits are consumed immature, fruits at this stage typically have different 

degrees of greenness, although some have yellow hue and even white color (Figure 1.2B). In 

mature fruit, however, there is a wide color variation ranging from white-green-yellow-orange-

brown (Figure 1.2C). Chlorophyll and carotene are the main two players influencing fruit color 

(Egea et al., 2010). Several genes that cause light green skin color were identified through 

16 

 
mutagenesis, and were found to be involved in different processes affecting chloroplast number or 

chlorophyll biosynthesis. Disrupting the expression of a recessive gene lgp, which encodes a 

homolog of ACCUMULATION AND REPLICATION OF CHLOROPLASTS 5 (ARC5), resulted in 

fewer but larger chloroplasts, probably due to interference with chloroplast division (Zhou et al., 

2015b). Another gene, CsYcf54 encoding an Ycf54-like protein localized in the chloroplast, was 

associated with a cyclase step in chlorophyll biosynthesis (Lun et al., 2016). Downregulating a 

transcription factor MYB36 from the CsMYB36 gene also led to decreased chlorophyll content in a 

yellow green peel mutant (ygp) (Hao et al., 2018). A major locus for pericarp color, 

CsPC1 (Pericarp color 1), encoding a GATA transcription factor (CsGATA1) that had higher 

expression in light green pericarp, was implicated to be involved in chlorophyll biosynthesis 

(Huang et al., 2022). White peel color in immature fruit with lower chlorophyll content and fewer 

and smaller chloroplast was affected by a two-component response regulator-like APRR2 (Liu et 

al., 2016; Jiao et al., 2017; Tang et al., 2018; Kishor et al., 2021a).  

Spine density 

Cucumber spines are multicellular non-glandular trichomes that consisted of an enlarged 

base and pointed stalk (Liu et al., 2022). Higher spine density is commonly found in North China 

type and Beit Alpha type cucumbers (Weng, 2021). The numerous spines (ns) locus located on 

chromosome 2 was identified by QTL mapping and GWAS analysis (Zhang et al., 2016; Xie et 

al., 2018; Liu et al., 2022). Overexpressing and knock out analyses suggested that ns was 

Csa2G264590, an auxin transporter (CsAUX1) that is specifically expressed in fruit peel (Xie et 

al., 2018; Liu et al., 2022). A non-functional ns haplotype which resulted in higher spine density 

was specifically selected in Eurasian cucumber accessions (Liu et al., 2022). Similarly, 

originating from mutagenesis in Chinese accessions with ultra-high fruit spine density phenotype 

17 

 
[e.g. Csgl3 (glabrous-3) and Tril (Trichome-less) (Cui et al., 2016; Bo et al., 2019; Du et al., 

2020)] were several allelic variations of an HD-Zip IV transcription factor essential for trichome 

formation in tomato (Yang et al., 2011). Spine initiation and differentiation is also associated with 

a WD-repeat homolog CsTTG1 (TRANSPARENT TESTA GLABRA1), which positively regulated 

the density of fruit bloom trichomes and spines (Chen et al., 2016; Guo et al., 2020). CsTTG1 also 

directly interacted with a class I homeodomain-leucine zipper gene, for which several mutants 

were identified including mict (micro-trichome)/csgl1 (glabrous-1)/tbh (tiny branched hair). 

Mutations of this gene led to tiny and stunted trichomes (Li et al., 2015; Zhao et al., 2015; Zhang 

et al., 2021b). Another factor, CsMYB6, a homolog of MYB6 in Arabidopsis encoding a MIXTA-

like MYB transcription factor, acts together with CsTRY , a homolog of Arabidopsis 

TRIPTYCHON, to negatively regulate and decrease trichome density (Yang et al., 2018). 

Netting 

Netted fruits are not often seen in commercial cucumber varieties, but can be observed in 

landraces or wild accessions, such as Sikkim type cucumber (Wang et al., 2021b; Weng, 2021). 

Microcracking on fruit surface occurs when cuticles on the outer cell walls fail to withstand the 

pressure of expansive growth (Knoche and Lang, 2017). These microcracks then may proceed to 

form fruit russeting, and netting (reticulation) (Petit et al., 2017; Zhang et al., 2022). Underneath 

the cracked surface is a layer of suberized wound-periderm in Sikkim type cucumber, which 

differs from cultivated cucumber, where the surface is composed of cutin and wax (Nomberg et 

al., 2022). A Rs (Russet skin) locus that conferred flaky skin that easily fell of the fruit was 

mapped to a 736 kb region on chromosome 1 (Wang et al., 2021b). Copy number variation of the 

Rs locus candidate gene, CsSH1/WIN1 (shine1/ WAX INDUCER1), which is also associated with 

cuticle thickness (Rett-Cadman et al., 2019), was positively correlated with netting intensity 

18 

 
(Zhang et al., 2022). Another dominant locus, H (Heavy netting), was located chromosome 5, but 

is yet to be cloned (Miao et al., 2011; Zhou et al., 2015a; Wang et al., 2021b). 

Genetic control of cucumber fruit internal traits 

Cucumber produces fleshy, pepo fruit that consist of three parts: exocarp (or called 

epicarp), which is the outer layer; mesocarp, the fleshy and crisp middle layer; and the endocarp, 

the inner pulp layer around the seeds, which is also known as seed cavity. The proportion of fruit 

flesh and seed cavity size is of commercial value; in general, thicker flesh and smaller seed cavity 

size are preferred by consumers. In addition to flesh thickness and seed cavity size, other internal 

characteristics such as flesh color and hollowness are also important traits that impact commercial 

value of a variety or cultivar.  

Carpel number 

Most cucumber accessions have three carpels, although some accessions or varieties have 

five carpels, such as XIS cucumber and True Lemon type cucumber. Carpel number in cucumber 

is controlled by a major gene (Cn) located on chromosome 1, where three carpels are dominant to 

five carpels in True Lemon cucumber (Li et al., 2016). Cn gene is a homolog of CLAVATA3 

(CsCLV3), which is a component of the well-characterized WUSCHEL (WUS)-

CLAVATA3 (CLV3) pathway that regulates shoot apical meristem maintenance and floral organ 

development (Li et al., 2016; Somssich et al., 2016; Che et al., 2020). CsCLV3 act as a negative 

regulator of carpel number and indirectly suppress CsWUS expression, while CsWUS acts as a 

promoter to increase variation in carpel number. Another gene, CsFUL1A, a FRUITFULL-

like MADS-box gene, can bind to the promoter of CsWUS and induce expression, and CsWUS can 

further bind to the promoter of CsCLV3 to activate its expression (Che et al., 2020). Additionally, 

auxin can also positively affect carpel number by interaction of AUXIN RESPONSE FACTOR 14 

19 

 
(CsARF14) with CsWUS (Che et al., 2020). 

Flesh thickness 

Flesh thickness represents the edible portion of a cucumber fruit. A QTL on chromosome 

2 (fft2.1) was identified from an F2 population crossed between a thick fruit flesh parent and a thin 

fruit flesh parent. From the QTL region of 0.19 Mb, a candidate gene Csa2M058670.1 was 

identified, which had higher expression in thick flesh line (Xu et al., 2015). In addition to low 

expression level of Csa2M058670.1, a 4-bp deletion in the promotor region was found in all four 

thin flesh lines, suggesting the gene activity was associated with fruit flesh phenotypes (Xu et al., 

2015). Csa2M058670.1 encodes a homolog of a SET domain protein-lysine methyltransferase 

(PKMT) (Raunser et al., 2009) that is associated with cell division and growth in Arabidopsis 

(Horvath et al., 2003). Csa2M058670 was specifically selected in a Xishuangbanna population, 

for which thinner fruit flesh is preferred (Lin et al., 2022). An additional six QTL were identified 

on chromosomes 1, 2, 3, and 6 from a RIL population derived from a North China type cucumber 

and a North European type cucumber (Yuan et al., 2008). Two bi-parental populations with 

different lines of Sikkim cucumber were also used to map various fruit quality QTL (Wang et al., 

2021b). Two flesh thickness QTL, fth3.1 and fth5.1, were identified on chromosomes 3 and 5, 

respectively, in both immature and mature fruits; fth3.1 was associated with increased flesh 

thickness while fth5.1 had the opposite effect.  

Seed cavity size  

Seed cavity size is highly correlated with fruit radial growth (Liu et al., 2020). 

Parthenocarpic fruits (i.e. without fertilization) that are widely grown for greenhouse production 

tend to have a greatly reduced seed cavity. Two QTL for seed cavity size at chromosomes 1 and 2 

were found from a population derived from a cross between parthenocarpic and non-

20 

 
parthenocarpic lines (Haaring and Huijbregts-doorduin, 2020). Yuan et al. (2008) identified seven 

QTL located on chromosomes 1, 2, 4, and 5 that explained 4-11.72% of phenotypic variation 

from the RIL population (S94 × S06). The QTL on chromosome 1 from both studies were located 

in similar location (Wang et al., 2020b). In a RIL population derived from a cultivated cucumber 

(CMCC) and an introgression line with C. hystrix background, two seed cavity size related QTL 

were identified on chromosome 5, and 6 (Wang et al., 2023b). 

Hollowness  

After double fertilization, a pistil comprised of 3-5 carpels develops into fruit and seeds of 

a cucumber. During the development, carpels fuse and the cell boundary between carpels become 

irregular and obscure (Zhou et al., 2022). Where carpels fail to fuse, the center becomes hollow. 

Hollowness is an undesirable trait that is not commonly seen in cultivated cucumber, especially in 

pickling cucumber, which may result in bloating effect during the brining process (Wilson and 

Baker, 1976). While environmental factors can affect hollowness formation, some cucumbers 

naturally form hollows, such as Sikkim type cucumber, where the size of hollowness was 

positively correlated with fruit diameter (Wang et al., 2021b). QTL mapping on populations with 

Sikkim cucumber background identified QTL on chromosome 1, 2,3, and 5, where mfh1.1 and 

mfh2.1 may increase the hollowness size while mfh3.1 decreases hollowness (Wang et al., 2021b). 

In Arabidopsis, two bHLH transcription factors, SPATULA (SPT) and ALCATRAZ (ALC), 

regulate differentiation of the transmitting tract, which connects the stigma style and ovary 

(Heisler et al., 2001; Groszmann et al., 2008; Pabón-Mora et al., 2014). The double knockout 

mutant Csspt Csalt in cucumber causes a hollow center in cucumber fruit, and reduces fertility 

(Cheng et al., 2022). In the mutant, lignin accumulation in the transmitting tract promoted the 

separation of carpels and resulted in hollowness formation (Cheng et al., 2022). Another study 

21 

 
identified a candidate gene, Csa1G630860 (CsALMT2), that encodes an aluminum-activated 

malate transporter protein ALMT2 associated with fruit hollowness (Zhou et al., 2022). CsALMT2 

was highly expressed in ovules and non-hollow fruit material at all fruit development stages 

compared to hollow fruit material (Zhou et al., 2022). A homolog in watermelon WMALMT-3 is 

involved in regulation of malic acid accumulation in vacuoles of pulp tissue during fruit 

development (Muhammad Jawad et al., 2020).  

Flesh color  

Typically, cucumber flesh color is white, with some ranging from green to yellow and 

orange. Green flesh can be found in C. hardwickii, and Xishuangbanna cucumber fruits are 

known for orange flesh (Qi et al., 1983; Renner, 2017). The yellow and orange are due to 

accumulation of carotene in mesocarp and endocarp, while green flesh results from chlorophyll 

accumulation (Li and Yuan, 2013). Orange flesh has been a breeding target due to the nutritional 

value of β-carotene content, which is a key precursor of vitamin A. Orange flesh is controlled by a 

single recessive gene, ore (Bo et al., 2012). A candidate gene was later identified for ore, 

Csa3G183920, which encodes B-carotene hydroxylase (CasBCH1) (Qi et al., 2013). Another 

orange flesh related gene, CsOr, was discovered from bi-parental populations from semi-white 

cucumber and an orange endocarp breeding line (Kishor et al., 2021b). The single recessive gene 

(CsaV3_6G040750) encodes a chaperone DnaJ protein (DnaJ) protein, which is associated with 

orange color in cauliflower and melon (Lu et al., 2006; Tzuri et al., 2015). A single recessive 

gene, yf, spanning 149 kb on chromosome 7 was identified from a yellow flesh and white flesh 

parents (Lu et al., 2015). Wang et al. (2023a) pinpointed another candidate gene, yellow flesh 2 

(Csyf2), that encodes an abscisic acid (ABA) 8’-hydroxylase. Downregulating Csyf2 expression 

led to lower levels of lutein and higher β-cryptoxanthin contents without changing the β-carotene 

22 

 
content (Wang et al., 2023a). Lastly, green flesh was found to be controlled by two QTL, 

qgf3.1 and qgf5.1, from a GWAS analysis (Bo et al., 2019).  

Objectives 

The second objective of this thesis was to morphologically characterize the cucumber core 

collection to provide publicly available photographic and phenotypic data that can be informative 

to breeders and researchers when choosing plant materials. I also sought to perform GWAS 

analyses to provide insights into the genetic basis of morphological diversity in the cucumber core 

collection. 

23 

 
 
 
WORKS CITED 

Abbo, S., Pinhasi van-Oss, R., Gopher, A., Saranga, Y., Ofner, I., and Peleg, Z. (2014). Plant 

domestication versus crop evolution: a conceptual framework for cereals and grain 
legumes. Trends in Plant Science 19, 351–360. doi: 10.1016/j.tplants.2013.12.002. 

Alzohairy, S. A., Hammerschmidt, R., and Hausbeck, M. K. (2020). Changes in Winter Squash 
Fruit Exocarp Structure Associated with Age-Related Resistance to Phytophthora 
capsici. Phytopathology 110, 447–455. doi: 10.1094/PHYTO-04-19-0128-R. 

Alzohairy, S. A., Hammerschmidt, R., and Hausbeck, M. K. (2021). Antifungal activity in winter 

squash fruit peel in relation to age related resistance to Phytophthora capsici. 
Physiological and Molecular Plant Pathology 114, 101603. doi: 
10.1016/j.pmpp.2021.101603. 

Ando, K., Carr, K. M., Colle, M., Mansfeld, B. N., and Grumet, R. (2015). Exocarp Properties 

and Transcriptomic Analysis of Cucumber (Cucumis sativus) Fruit Expressing Age-
Related Resistance to Phytophthora capsici. PLOS ONE 10, e0142133. doi: 
10.1371/journal.pone.0142133. 

Ando, K., Carr, K. M., and Grumet, R. (2012). Transcriptome analyses of early cucumber fruit 

growth identifies distinct gene modules associated with phases of development. BMC 
Genomics 13, 518. doi: 10.1186/1471-2164-13-518. 

Ando, K., and Grumet, R. (2006). Evaluation of Altered Cucumber Plant Architecture as a 

Means to Reduce Phytophthora capsici Disease Incidence on Cucumber Fruit. J. Amer. 
Soc. Hort. Sci. 131, 491–498. 

Ando, K., Hammar, S., and Grumet, R. (2009). Age-related resistance of diverse cucurbit fruit to 

infection by Phytophthora capsici. J. Amer. Soc. Hort. Sci. 134, 176–182. 

Babadoost, M. (2005). Phytophthora blight of cucurbits. The Plant Health Instructor. doi: 

10.1094/PHI-I-2005-0429-01. 

Babadoost, M., and Pavon, C. (2013). Survival of oospores of Phytophthora capsici in soil. Plant 

Disease 97, 1478–1483. doi: 10.1094/PDIS-12-12-1123-RE. 

Bo, K., Ma, Z., Chen, J., and Weng, Y. (2015). Molecular mapping reveals structural 

rearrangements and quantitative trait loci underlying traits with local adaptation in semi-
wild Xishuangbanna cucumber (Cucumis sativus L. var. xishuangbannanesis Qi et 
Yuan). Theor Appl Genet 128, 25–39. doi: 10.1007/s00122-014-2410-z. 

Bo, K., Song, H., Shen, J., Qian, C., Staub, J. E., Simon, P. W., et al. (2012). Inheritance and 
mapping of the ore gene controlling the quantity of β-carotene in cucumber (Cucumis 
sativus L.) endocarp. Mol Breeding 30, 335–344. doi: 10.1007/s11032-011-9624-4. 

24 

 
Bo, K., Wei, S., Wang, W., Miao, H., Dong, S., Zhang, S., et al. (2019). QTL mapping and 

genome-wide association study reveal two novel loci associated with green flesh color in 
cucumber. BMC Plant Biology 19, 243. doi: 10.1186/s12870-019-1835-6. 

Café-Filho, A. C., and Ristaino, J. B. (2008). Fitness of Isolates of Phytophthora capsici 

Resistant to Mefenoxam from Squash and Pepper Fields in North Carolina. Plant Disease 
92, 1439–1443. doi: 10.1094/PDIS-92-10-1439. 

Che, G., Gu, R., Zhao, J., Liu, X., Song, X., Zi, H., et al. (2020). Gene regulatory network 

controlling carpel number variation in cucumber. Development 147, dev184788. doi: 
10.1242/dev.184788. 

Che, G., Pan, Y., Liu, X., Li, M., Zhao, J., Yan, S., et al. (2023). Natural variation in CRABS 

CLAW contributes to fruit length divergence in cucumber. The Plant Cell 35, 738–755. 
doi: 10.1093/plcell/koac335. 

Che, G., and Zhang, X. (2019). Molecular basis of cucumber fruit domestication. Current 

Opinion in Plant Biology 47, 38–46. doi: 10.1016/j.pbi.2018.08.006. 

Chen, C., Yin, S., Liu, X., Liu, B., Yang, S., Xue, S., et al. (2016). The WD-Repeat Protein 

CsTTG1 Regulates Fruit Wart Formation through Interaction with the Homeodomain-
Leucine Zipper I Protein Mict1. Plant Physiol 171, 1156–1168. doi: 
10.1104/pp.16.00112. 

Cheng, Z., Song, X., Liu, X., Yan, S., Song, W., Wang, Z., et al. (2022). SPATULA and 

ALCATRAZ confer female sterility and fruit cavity via mediating pistil development in 
cucumber. Plant Physiology 189, 1553–1569. doi: 10.1093/plphys/kiac158. 

Colle, M., Straley, E. N., Makela, S. B., Hammar, S. A., and Grumet, R. (2014). Screening the 

cucumber plant introduction collection for young fruit resistance to Phytophthora capsici. 
HortScience 49, 244–249. 

Cuevas, H. E., Song, H., Staub, J. E., and Simon, P. W. (2010). Inheritance of beta-carotene-

associated flesh color in cucumber (Cucumis sativus L.) fruit. Euphytica 171, 301–311. 
doi: 10.1007/s10681-009-0017-2. 

Cui, J.-Y., Miao, H., Ding, L.-H., Wehner, T. C., Liu, P.-N., Wang, Y., et al. (2016). A New 
Glabrous Gene (csgl3) Identified in Trichome Development in Cucumber (Cucumis 
sativus L.). PLOS ONE 11, e0148422. doi: 10.1371/journal.pone.0148422. 

Davey, J. F., Gregory, N. F., Mulrooney, R. P., Evans, T. A., and Carroll, R. B. (2008). First 

Report of Mefenoxam-Resistant Isolates of Phytophthora capsici from Lima Bean Pods 
in the Mid-Atlantic Region. Plant Disease 92, 656–656. doi: 10.1094/PDIS-92-4-0656A. 

Deakin, J. R., Bohn, G. W., and Whitaker, T. W. (1971). Interspecific Hybridization in Cucumis. 

Economic Botany 25, 195–211. 

25 

 
 
Donahoo, R. S., Turechek, W. W., Thies, J. A., & Kousik, C. S. (2013). Potential Sources of 
Resistance in U.S. Cucumis melo PIs to Crown Rot Caused by Phytophthora capsici. 
HortScience horts, 48(2), 164-170. 

Du, H., Wang, G., Pan, J., Chen, Y., Xiao, T., Zhang, L., et al. (2020). The HD-ZIP IV 

transcription factor Tril regulates fruit spine density through gene dosage effects in 
cucumber. Journal of Experimental Botany 71, 6297–6310. doi: 10.1093/jxb/eraa344. 

Dunn, A. R., Milgroom, M. G., Meitz, J. C., McLeod, A., Fry, W. E., McGrath, M. T., et al. 

(2010). Population Structure and Resistance to Mefenoxam of Phytophthora capsici in 
New York State. Plant Disease 94, 1461–1468. doi: 10.1094/PDIS-03-10-0221. 

Egea, I., Barsan, C., Bian, W., Purgatto, E., Latché, A., Chervin, C., et al. (2010). Chromoplast 
differentiation: current status and perspectives. Plant Cell Physiol 51, 1601–1611. doi: 
10.1093/pcp/pcq136. 

Gevens,  a J., Ando, K., Lamour, K. H., Grumet, R., and Hausbeck, M. K. (2006). A Detached 
Cucumber Fruit Method to Screen for Resistance to Phytophthora capsici and Effect of 
Fruit Age on Susceptibility to Infection. Plant Disease 90, 1276–1282. doi: 10.1094/PD-
90-1276. 

Granke, L. L., Quesada-Ocampo, L., Lamour, K., and Hausbeck, M. K. (2012). Advances in 

research on Phytophthora capsici on vegetable crops in the United States. Plant Disease 
96, 1588–1600. doi: 10.1094/PDIS-02-12-0211-FE. 

Groszmann, M., Paicu, T., and Smyth, D. R. (2008). Functional domains of SPATULA, a bHLH 
transcription factor involved in carpel and fruit development in Arabidopsis. Plant J 55, 
40–52. doi: 10.1111/j.1365-313X.2008.03469.x. 

Grumet, R., and Colle, M. (2017). Cucumber (Cucumis sativus) Breeding Line with Young Fruit 
Resistance to Infection by Phytophthora capsici. HortScience 52, 922–924. doi: 
10.21273/HORTSCI11423-16. 

Grumet, R., Fei, Z., Levi, A., Mazourek, M., McCreight, J. D., Schultheis, J., et al. (2020). The 
CucCAP project: leveraging applied genomics to improve disease resistance in cucurbit 
crops. Acta Hortic., 91–104. doi: 10.17660/ActaHortic.2020.1294.12. 

Grumet, R., Lin, Y.-C., Rett-Cadman, S., and Malik, A. (2023). Morphological and Genetic 
Diversity of Cucumber (Cucumis sativus L.) Fruit Development. Plants 12, 23. doi: 
10.3390/plants12010023. 

Guo, P., Chang, H., Li, Q., Wang, L., Ren, Z., Ren, H., et al. (2020). Transcriptome profiling 

reveals genes involved in spine development during CsTTG1-regulated pathway in 
cucumber (Cucumis sativus L.). Plant Science 291, 110354. doi: 
10.1016/j.plantsci.2019.110354. 

26 

 
Haaring, C., and Huijbregts-doorduin, L. J. (2020). Genetic basis for cucumber fruit having small 
seed cavity. Available at: https://patents.google.com/patent/US10736288B2/en [Accessed 
October 10, 2023]. 

Hao, N., Du, Y., Li, H., Wang, C., Wang, C., Gong, S., et al. (2018). CsMYB36 is involved in 

the formation of yellow green peel in cucumber (Cucumis sativus L.). Theor Appl Genet 
131, 1659–1669. doi: 10.1007/s00122-018-3105-7. 

Harlan, J. R. ed. (1992). Crops & Man. Madison, WI, USA: American Society of Agronomy, 

Crop Science Society of America doi: 10.2135/1992.cropsandman. 

Hausbeck, M. K., and Lamour, K. H. (2004). Phytophthora capsici on vegetable crops: research 

progress and management challenges. Plant Disease 88, 1292–1303. doi: 
10.1094/PDIS.2004.88.12.1292. 

Heisler, M. G., Atkinson, A., Bylstra, Y. H., Walsh, R., and Smyth, D. R. (2001). SPATULA, a 

gene that controls development of carpel margin tissues in Arabidopsis, encodes a bHLH 
protein. Development 128, 1089–1098. doi: 10.1242/dev.128.7.1089. 

Hooker, J. D. (1876). Cucumis sativus var. sikkimensis collected in the Himalayan Mountains. 

Bot. Mag 32, 1067. 

Horvath, D. P., Schaffer, R., West, M., and Wisman, E. (2003). Arabidopsis microarrays identify 

conserved and differentially expressed genes involved in shoot growth and development 
from distantly related plant species. The Plant Journal 34, 125–134. doi: 10.1046/j.1365-
313X.2003.01706.x. 

Huang, H., Yang, Q., Zhang, L., Kong, W., Wang, H., Wei, A., et al. (2022). Genome-wide 

association analysis reveals a novel QTL CsPC1 for pericarp color in cucumber. BMC 
Genomics 23, 383. doi: 10.1186/s12864-022-08606-5. 

Huang, S., Li, R., Zhang, Z., Li, L., Gu, X., Fan, W., et al. (2009). The genome of the cucumber, 

Cucumis sativus L. Nature Genetics 41, 1275–1281. doi: 10.1038/ng.475. 

Jiao, J., Liu, H., Liu, J., Cui, M., Xu, J., Meng, H., et al. (2017). Identification and functional 

characterization of APRR2 controlling green immature fruit color in cucumber (Cucumis 
sativus L.). Plant Growth Regul 83, 233–243. doi: 10.1007/s10725-017-0304-1. 

Kishor, D. S., Alavilli, H., Lee, S.-C., Kim, J.-G., and Song, K. (2021a). Development of SNP 
Markers for White Immature Fruit Skin Color in Cucumber (Cucumis sativus L.) Using 
QTL-seq and Marker Analyses. Plants 10, 2341. doi: 10.3390/plants10112341. 

Kishor, D. S., Lee, H.-Y., Alavilli, H., You, C.-R., Kim, J.-G., Lee, S.-Y., et al. (2021b). 

Identification of an Allelic Variant of the CsOr Gene Controlling Fruit Endocarp Color in 
Cucumber (Cucumis sativus L.) Using Genotyping-By-Sequencing (GBS) and Whole-
Genome Sequencing. Frontiers in Plant Science 12. doi: 10.3389/fpls.2021.802864. 

27 

 
Knoche, M., and Lang, A. (2017). Ongoing Growth Challenges Fruit Skin Integrity. Critical 
Reviews in Plant Sciences 36, 190–215. doi: 10.1080/07352689.2017.1369333. 

Kousik, C. S., Donahoo, R. S., and Hassell, R. (2012). Resistance in watermelon rootstocks to 

crown rot caused by Phytophthora capsici. Crop Protection 39, 18–25. doi: 
10.1016/j.cropro.2012.04.004. 

Kousik, C. S., Ikerd, J. L., Wechter, W. P., Branham, S., and Turechek, W. (2022). Broad 

Resistance to Post-Harvest Fruit Rot in USVL Watermelon Germplasm Lines to Isolates 
of Phytophthora capsici Across the United States. Plant Disease 106, 711–719. doi: 
10.1094/PDIS-11-20-2480-RE. 

Kousik, C. S., Ikerd, J., Mandal, M., Adkins, S., and Turechek, W. W. (2023). USVL531-MDR: 

Watermelon Germplasm Line with Broad Resistance to Powdery Mildew and 
Phytophthora Fruit Rot. HortScience 58, 475–479. doi: 10.21273/HORTSCI16907-22. 

Kousik, C. (Shaker) S., Ji, P., Egel, D. S., and Quesada-Ocampo, L. M. (2017). Fungicide 

Rotation Programs for Managing Phytophthora Fruit Rot of Watermelon in Southeastern 
United States. Plant Health Progress 18, 28–34. doi: 10.1094/PHP-RS-16-0059. 

Krasnow, C. S., Hammerschmidt, R., and Hausbeck, M. K. (2017). Characteristics of Resistance 

to Phytophthora Root and Crown Rot in Cucurbita pepo. Plant Disease 101, 659–665. 
doi: 10.1094/PDIS-06-16-0867-RE. 

Krasnow, C. S., Naegele, R. P., and Hausbeck, M. K. (2014). Evaluation of fruit rot resistance in 

cucurbita germplasm resistant to Phytophthora capsici crown rot. HortScience 49, 285–
288. 

Lamour, K. (2013). Phytophthora: a global perspective. , ed. K. Lamour Wallingford: CABI doi: 

10.1079/9781780640938.0000. 

Lamour, K. H., and Hausbeck, M. K. (2000). Mefenoxam Insensitivity and the Sexual Stage of 

Phytophthora capsici in Michigan Cucurbit Fields. Phytopathology 90, 396–400. doi: 
10.1094/PHYTO.2000.90.4.396. 

Lamour, K. H., Stam, R., Jupe, J., and Huitema, E. (2012). The oomycete broad-host-range 
pathogen Phytophthora capsici. Molecular Plant Pathology 13, 329–337. doi: 
10.1111/j.1364-3703.2011.00754.x. 

LaPlant, K. E., Vogel, G., Reeves, E., Smart, C. D., and Mazourek, M. (2020). Performance and 

Resistance to Phytophthora Crown and Root Rot in Squash Lines. HortTechnology 30, 
608–618. doi: 10.21273/HORTTECH04636-20. 

Li, L., and Yuan, H. (2013). Chromoplast biogenesis and carotenoid accumulation. Arch 

Biochem Biophys 539, 102–109. doi: 10.1016/j.abb.2013.07.002. 

Li, Q., Cao, C., Zhang, C., Zheng, S., Wang, Z., Wang, L., et al. (2015). The identification of 

Cucumis sativus Glabrous 1 (CsGL1) required for the formation of trichomes uncovers a 

28 

 
novel function for the homeodomain-leucine zipper I gene. Journal of Experimental 
Botany 66, 2515–2526. doi: 10.1093/jxb/erv046. 

Li, Q., Li, H., Huang, W., Xu, Y., Zhou, Q., Wang, S., et al. (2019). A chromosome-scale 
genome assembly of cucumber (Cucumis sativus L.). GigaScience 8, giz072. doi: 
10.1093/gigascience/giz072. 

Li, S., Pan, Y., Wen, C., Li, Y., Liu, X., Zhang, X., et al. (2016). Integrated analysis in bi-

parental and natural populations reveals CsCLAVATA3 (CsCLV3) underlying carpel 
number variations in cucumber. Theor Appl Genet 129, 1007–1022. doi: 10.1007/s00122-
016-2679-1. 

Li, S., Wang, C., Zhou, X., Liu, D., Liu, C., Luan, J., et al. (2020). The curvature of cucumber 

fruits is associated with spatial variation in auxin accumulation and expression of a 
YUCCA biosynthesis gene. Hortic Res 7, 1–12. doi: 10.1038/s41438-020-00354-5. 

Lin, X., Zhang, N., Song, H., Lin, K., and Pang, E. (2022). Population-specific, recent positive 

selection signatures in cultivated Cucumis sativus L. (cucumber). G3 (Bethesda) 12, 
jkac119. doi: 10.1093/g3journal/jkac119. 

Liu, H., Jiao, J., Liang, X., Liu, J., Meng, H., Chen, S., et al. (2016). Map-based cloning, 

identification and characterization of the w gene controlling white immature fruit color in 
cucumber (Cucumis sativus L.). Theor Appl Genet 129, 1247–1256. doi: 10.1007/s00122-
016-2700-8. 

Liu, X., Pan, Y., Liu, C., Ding, Y., Wang, X., Cheng, Z., et al. (2020). Cucumber Fruit Size and 
Shape Variations Explored from the Aspects of Morphology, Histology, and Endogenous 
Hormones. Plants 9, 772. doi: 10.3390/plants9060772. 

Liu, X., Yang, X., Xie, Q., Miao, H., Bo, K., Dong, S., et al. (2022). NS encodes an auxin 

transporter that regulates the ‘numerous spines’ trait in cucumber (Cucumis sativus) fruit. 
The Plant Journal 110, 325–336. doi: 10.1111/tpj.15710. 

Lu, H. W., Miao, H., Tian, G. L., Wehner, T. C., Gu, X. F., and Zhang, S. P. (2015). Molecular 

mapping and candidate gene analysis for yellow fruit flesh in cucumber. Molecular 
Breeding 35, 64. doi: 10.1007/s11032-015-0263-z. 

Lu, S., Van Eck, J., Zhou, X., Lopez, A. B., O’Halloran, D. M., Cosman, K. M., et al. (2006). 
The cauliflower Or gene encodes a DnaJ cysteine-rich domain-containing protein that 
mediates high levels of beta-carotene accumulation. Plant Cell 18, 3594–3605. doi: 
10.1105/tpc.106.046417. 

Lun, Y., Wang, X., Zhang, C., Yang, L., Gao, D., Chen, H., et al. (2016). A CsYcf54 variant 
conferring light green coloration in cucumber. Euphytica 208, 509–517. doi: 
10.1007/s10681-015-1592-z. 

29 

 
Lv, J., Qi, J., Shi, Q., Shen, D., Zhang, S., Shao, G., et al. (2012). Genetic Diversity and 

Population Structure of Cucumber (Cucumis sativus L.). PLoS ONE 7, e46919. doi: 
10.1371/journal.pone.0046919. 

Mansfeld, B. N., Colle, M., Kang, Y., Jones, A. D., and Grumet, R. (2017). Transcriptomic and 
metabolomic analyses of cucumber fruit peels reveal a developmental increase in 
terpenoid glycosides associated with age-related resistance to Phytophthora capsici. 
Hortic Res 4, 1–11. doi: 10.1038/hortres.2017.22. 

Mansfeld, B. N., Colle, M., Zhang, C., Lin, Y.-C., and Grumet, R. (2020). Developmentally 
regulated activation of defense allows for rapid inhibition of infection in age-related 
resistance to Phytophthora capsici in cucumber fruit. BMC Genomics 21, 628. doi: 
10.1186/s12864-020-07040-9. 

McCreight, J. D., Staub, J. E., Wehner, T. C., and Dhillon, N. P. S. (2013). Gone Global: 

Familiar and Exotic Cucurbits Have Asian Origins. HortScience 48, 1078–1089. doi: 
10.21273/HORTSCI.48.9.1078. 

Miao, H., Zhang, S., Wang, X., Zhang, Z., Li, M., Mu, S., et al. (2011). A linkage map of 

cultivated cucumber (Cucumis sativus L.) with 248 microsatellite marker loci and seven 
genes for horticulturally important traits. Euphytica 182, 167–176. doi: 10.1007/s10681-
011-0410-5. 

Michael, V. N., Fu, Y., and Meru, G. (2019). Inheritance of Resistance to Phytophthora Crown 
Rot in Cucurbita pepo. HortScience 54, 1156–1158. doi: 10.21273/HORTSCI14021-19. 

Michael, V. N., Fu, Y., Shrestha, S., and Meru, G. (2021). A Novel QTL for Resistance to 
Phytophthora Crown Rot in Squash. Plants 10, 2115. doi: 10.3390/plants10102115. 

Muhammad Jawad, U., Gao, L., Gebremeskel, H., Safdar, L. B., Yuan, P., Zhao, S., et al. (2020). 

Expression pattern of sugars and organic acids regulatory genes during watermelon fruit 
development. Scientia Horticulturae 265, 109102. doi: 10.1016/j.scienta.2019.109102. 

Naegele, R. P., and Wehner, T. C. (2016). Genetic Resources of Cucumber. Plant Genetics and 

Genomics: Crops and Models, 1–26. doi: 10.1007/7397_2016_15. 

Navazio, J. P., and Simon, P. W. (2001). Diallel Analysis of High Carotenoid Content in 

Cucumbers. Journal of the American Society for Horticultural Science 126, 100–104. 
doi: 10.21273/JASHS.126.1.100. 

Niu, H., Liu, X., Tong, C., Wang, H., Li, S., Lu, L., et al. (2018). The WUSCHEL-related 

homeobox1 gene of cucumber regulates reproductive organ development. Journal of 
Experimental Botany 69, 5373–5387. doi: 10.1093/jxb/ery329. 

Nomberg, G., Marinov, O., Karavani, E., Manasherova, E., Zelinger, E., Yarden, O., et al. 

(2022). Cucumber fruit skin reticulation affects post-harvest traits. Postharvest Biology 
and Technology 194, 112071. doi: 10.1016/j.postharvbio.2022.112071. 

30 

 
Ogundiwin, E. A., Berke, T. F., Massoudi, M., Black, L. L., Huestis, G., Choi, D., et al. (2005). 

Construction of 2 intraspecific linkage maps and identification of resistance QTLs for 
Phytophthora capsici root-rot and foliar-blight diseases of pepper (Capsicum annuum L.). 
Genome 48, 698–711. doi: 10.1139/g05-028. 

Osipowski, P., Pawełkowicz, M., Wojcieszek, M., Skarzyńska, A., Przybecki, Z., and Pląder, W. 
(2020). A high-quality cucumber genome assembly enhances computational comparative 
genomics. Mol Genet Genomics 295, 177–193. doi: 10.1007/s00438-019-01614-3. 

Pabón-Mora, N., Wong, G. K.-S., and Ambrose, B. A. (2014). Evolution of fruit development 
genes in flowering plants. Frontiers in Plant Science 5. doi: 10.3389/fpls.2014.00300. 

Padley, Les D., J., Kabelka, E. A., and Roberts, P. D. (2009). Inheritance of resistance to crown 

rot caused by Phytophthora capsici in Cucurbita. HortScience 44, 211–213. 

Pan, Y., Chen, B., Qiao, L., Chen, F., Zhao, J., Cheng, Z., et al. (2022). Phenotypic 

Characterization and Fine Mapping of a Major-Effect Fruit Shape QTL FS5.2 in 
Cucumber, Cucumis sativus L., with Near-Isogenic Line-Derived Segregating 
Populations. International Journal of Molecular Sciences 23, 13384. doi: 
10.3390/ijms232113384. 

Pan, Y., Liang, X., Gao, M., Liu, H., Meng, H., Weng, Y., et al. (2017). Round fruit shape in 

WI7239 cucumber is controlled by two interacting quantitative trait loci with one 
putatively encoding a tomato SUN homolog. Theor Appl Genet 130, 573–586. doi: 
10.1007/s00122-016-2836-6. 

Pan, Y., Wang, Y., McGregor, C., Liu, S., Luan, F., Gao, M., et al. (2020). Genetic architecture 

of fruit size and shape variation in cucurbits: a comparative perspective. Theor Appl 
Genet 133, 1–21. doi: 10.1007/s00122-019-03481-3. 

Paris, H. S., Daunay, M.-C., and Janick, J. (2012). Occidental diffusion of cucumber (Cucumis 
sativus) 500-1300 CE: two routes to Europe. Annals of Botany 109, 117–126. doi: 
10.1093/aob/mcr281. 

Parra, G., and Ristaino, J. B. (2001). Resistance to Mefenoxam and Metalaxyl Among Field 
Isolates of Phytophthora capsici Causing Phytophthora Blight of Bell Pepper. Plant 
Disease 85, 1069–1075. doi: 10.1094/PDIS.2001.85.10.1069. 

Petit, J., Bres, C., Mauxion, J.-P., Bakan, B., and Rothan, C. (2017). Breeding for cuticle-

associated traits in crop species: traits, targets, and strategies. Journal of Experimental 
Botany 68, 5369–5387. doi: 10.1093/jxb/erx341. 

Pontes, N. de C., Aguiar, F. M., Boiteux, L. S., Lima, M. L. P., Oliveira, V. R., Caf? Filho, A. 

C., et al. (2014). Identification of sources of seedling resistance to Phytophthora capsici 
in Cucumis melo. Tropical Plant Pathology 39, 74–81. doi: 10.1590/S1982-
56762014000100009. 

31 

 
Qi, C. Z., Yuan, Z. Z., and Li, Y. X. (1983). A new type of cucumber. Cucumis sativus L. var. 

xishuang-bannanesis. 

Qi, J., Liu, X., Shen, D., Miao, H., Xie, B., Li, X., et al. (2013). A genomic variation map 

provides insights into the genetic basis of cucumber domestication and diversity. Nature 
Genetics 45, 1510–1515. doi: 10.1038/ng.2801. 

Quesada-Ocampo, L. M., and Hausbeck, M. K. (2010). Resistance in Tomato and Wild Relatives 
to Crown and Root Rot Caused by Phytophthora capsici. Phytopathology 100, 619–627. 
doi: 10.1094/PHYTO-100-6-0619. 

Quesada-Ocampo, L. M., Parada-Rojas, C. H., Hansen, Z., Vogel, G., Smart, C., Hausbeck, M. 

K., et al. (2023). Phytophthora capsici: Recent Progress on Fundamental Biology and 
Disease Management 100 Years After Its Description. Annual Review of Phytopathology 
61, null. doi: 10.1146/annurev-phyto-021622-103801. 

Quesada-Ocampo, L. M., Vargas, A. M., Naegele, R. P., Francis, D. M., and Hausbeck, M. K. 

(2016). Resistance to Crown and Root Rot Caused by Phytophthora capsici in a Tomato 
Advanced Backcross of Solanum habrochaites and Solanum lycopersicum. Plant Disease 
100, 829–835. doi: 10.1094/PDIS-08-15-0888-RE. 

Ramos, A., Michael, V., and Meru, G. (2020). QTL-seq for identification of loci associated with 

resistance to Phytophthora crown rot in squash. Sci Rep 10, 5326. doi: 10.1038/s41598-
020-62228-z. 

Raunser, S., Magnani, R., Huang, Z., Houtz, R. L., Trievel, R. C., Penczek, P. A., et al. (2009). 

Rubisco in complex with Rubisco large subunit methyltransferase. Proc Natl Acad Sci U 
S A 106, 3160–3165. doi: 10.1073/pnas.0810563106. 

Renner, S. S. (2017). A valid name for the Xishuangbanna gourd, a cucumber with carotene-rich 

fruits. PhytoKeys 85, 87–94. doi: 10.3897/phytokeys.85.17371. 

Renner, S. S., and Schaefer, H. (2016). Phylogeny and Evolution of the Cucurbitaceae. Plant 
Genetics and Genomics: Crops and Models, 1–26. doi: 10.1007/7397_2016_14. 

Rett-Cadman, S., Colle, M., Mansfeld, B., Barry, C. S., Wang, Y., Weng, Y., et al. (2019). QTL 
and Transcriptomic Analyses Implicate Cuticle Transcription Factor SHINE as a Source 
of Natural Variation for Epidermal Traits in Cucumber Fruit. Frontiers in Plant Science 
10. Available at: https://www.frontiersin.org/articles/10.3389/fpls.2019.01536. 

Savory, E. A., Granke, L. L., Quesada-Ocampo, L. M., Varbanova, M., Hausbeck, M. K., and 

Day, B. (2011). The cucurbit downy mildew pathogen Pseudoperonospora cubensis. 
Molecular Plant Pathology 12, 217–226. doi: 10.1111/j.1364-3703.2010.00670.x. 

Sebastian, P., Schaefer, H., Telford, I. R. H., and Renner, S. S. (2010). Cucumber (Cucumis 

sativus) and melon (C. melo) have numerous wild relatives in Asia and Australia, and the 
sister species of melon is from Australia. Proceedings of the National Academy of 
Sciences 107, 14269–14273. doi: 10.1073/pnas.1005338107. 

32 

 
Sharma, B. D., and Hore, D. K. (1996). Indian cucumber germplasm and challenges ahead. 

Genetic Resources and Crop Evolution 43, 7–12. 

Somssich, M., Je, B. I., Simon, R., and Jackson, D. (2016). CLAVATA-WUSCHEL signaling in 

the shoot meristem. Development 143, 3238–3248. doi: 10.1242/dev.133645. 

Staub, J. E., Robbins, M. D., and Wehner, T. C. (2008). Cucumber. Vegetables I: Asteraceae, 
Brassicaceae, Chenopodicaceae, and Cucurbitaceae (New York, NY: Springer New 
York), 241–282. doi: 10.1007/978-0-387-30443-4_8. 

Sy, O., Bosland, P. W., and Steiner, R. (2005). Inheritance of Phytophthora stem blight 

resistance as compared to Phytophthora root rot and Phytophthora foliar blight resistance 
in Capsicum annuum L. Journal of the American Society for Horticultural Science 130, 
75–78. 

Tang, H.-Y., Dong, X., Wang, J.-K., Xia, J.-H., Xie, F., Zhang, Y., et al. (2018). Fine Mapping 

and Candidate Gene Prediction for White Immature Fruit Skin in Cucumber (Cucumis 
sativus L.). International Journal of Molecular Sciences 19, 1493. doi: 
10.3390/ijms19051493. 

Tzuri, G., Zhou, X., Chayut, N., Yuan, H., Portnoy, V., Meir, A., et al. (2015). A ‘golden’ SNP 
in CmOr governs the fruit flesh color of melon (Cucumis melo). The Plant Journal 82, 
267–279. doi: 10.1111/tpj.12814. 

Uchneat, M. S., and Wehner, T. C. (1998). Resistance to Belly Rot in Cucumber Identified 
through Field and Detached-fruit Evaluations. Journal of the American Society for 
Horticultural Science 123, 78–84. 

Vogel, G., LaPlant, K. E., Mazourek, M., Gore, M. A., and Smart, C. D. (2021). A combined 

BSA-Seq and linkage mapping approach identifies genomic regions associated with 
Phytophthora root and crown rot resistance in squash. Theoretical and Applied Genetics. 
doi: 10.1007/s00122-020-03747-1. 

Walker, S. J., and Bosland, P. W. (1999). Inheritance of Phytophthora root rot and foliar blight 
resistance in pepper. Journal of the American Society for Horticultural Science v. 124. 

Walters, S. A., Wehner, T. C., and Barker, K. R. (1996). NC-42 and NC-43: Root-Knot 
Nematode–Resistant Cucumber Germplasm. HortSci 31, 1246–1247. doi: 
10.21273/HORTSCI.31.7.1246. 

Wang, H., Sun, J., Yang, F., Weng, Y., Chen, P., Du, S., et al. (2021a). CsKTN1 for a katanin 
p60 subunit is associated with the regulation of fruit elongation in cucumber (Cucumis 
sativus L.). Theor Appl Genet 134, 2429–2441. doi: 10.1007/s00122-021-03833-y. 

Wang, L., Cao, C., Zheng, S., Zhang, H., Liu, P., Ge, Q., et al. (2017). Transcriptomic analysis 
of short-fruit 1 (sf1) reveals new insights into the variation of fruit-related traits in 
Cucumis sativus. Sci Rep 7, 2950. doi: 10.1038/s41598-017-02932-5. 

33 

 
Wang, P., Xu, X., Zhao, G., He, Y., Hou, C., Kong, W., et al. (2020a). Genetic mapping and 

candidate gene analysis for melon resistance to Phytophthora capsici. Scientific Reports 
10, 20456. doi: 10.1038/s41598-020-77600-2. 

Wang, X., Bao, K., Reddy, U. K., Bai, Y., Hammar, S. A., Jiao, C., et al. (2018). The USDA 

cucumber (Cucumis sativus L.) collection: genetic diversity, population structure, 
genome-wide association studies, and core collection development. Hortic Res 5, 1–13. 
doi: 10.1038/s41438-018-0080-8. 

Wang, X., Jin, B., Yan, W., Wang, J., Xu, J., Cai, C., et al. (2023a). Cucumber abscisic acid 8’-
hydroxylase Csyf2 regulates yellow flesh by modulating carotenoid biosynthesis. Plant 
Physiol 193, 1001–1015. doi: 10.1093/plphys/kiad383. 

Wang, Y., Bo, K., Gu, X., Pan, J., Li, Y., Chen, J., et al. (2020b). Molecularly tagged genes and 
quantitative trait loci in cucumber with recommendations for QTL nomenclature. 
Horticulture Research 7, 3. doi: 10.1038/s41438-019-0226-3. 

Wang, Y., Fang, Y., Ning, S., Xia, L., Zhan, J., Yang, Z., et al. (2023b). QTL Mapping for 

Ovary- and Fruit-Related Traits in Cucumis sativus-C. hystrix Introgression Line IL52. 
Genes 14, 1133. doi: 10.3390/genes14061133. 

Wang, Y., Jiang, B., Dymerski, R., Xu, X., and Weng, Y. (2021b). Quantitative trait loci for 

horticulturally important traits defining the Sikkim cucumber, Cucumis sativus var. 
sikkimensis. Theor Appl Genet 134, 229–247. doi: 10.1007/s00122-020-03693-y. 

Weng, Y. (2021). “Cucumis sativus Chromosome Evolution, Domestication, and Genetic 
Diversity: Implications for Cucumber Breeding,” in Plant Breeding Reviews, ed. I. 
Goldman (Wiley), 79–111. doi: 10.1002/9781119717003.ch4. 

Weng, Y., Colle, M., Wang, Y., Yang, L., Rubinstein, M., Sherman, A., et al. (2015). QTL 

mapping in multiple populations and development stages reveals dynamic quantitative 
trait loci for fruit size in cucumbers of different market classes. Theoretical and Applied 
Genetics 128, 1747–1763. doi: 10.1007/s00122-015-2544-7. 

Weston, L. A., Geneve, R. L., and Staub, J. E. (1992). Seed dormancy in Cucumis sativus var. 
hardwickii (Royle) Alef. Scientia Horticulturae 50, 35–46. doi: 10.1016/S0304-
4238(05)80006-8. 

Wilson, J. E., and Baker, L. R. (1976). Inheritance of Carpel Separation in Mature Fruits of 

Pickling Cucumbers. Journal of the American Society for Horticultural Science 101, 66–
69. doi: 10.21273/JASHS.101.1.66. 

Wu, S., Zhang, B., Keyhaninejad, N., Rodríguez, G. R., Kim, H. J., Chakrabarti, M., et al. 

(2018). A common genetic mechanism underlies morphological diversity in fruits and 
other plant organs. Nat Commun 9, 4734. doi: 10.1038/s41467-018-07216-8. 

Xie, Q., Liu, P., Shi, L., Miao, H., Bo, K., Wang, Y., et al. (2018). Combined fine mapping, 

genetic diversity, and transcriptome profiling reveals that the auxin transporter gene ns 

34 

 
plays an important role in cucumber fruit spine development. Theor Appl Genet 131, 
1239–1252. doi: 10.1007/s00122-018-3074-x. 

Xie, Y., Liu, X., Sun, C., Song, X., Li, X., Cui, H., et al. (2023). CsTRM5 regulates fruit shape 

via mediating cell division direction and cell expansion in cucumber. Horticulture 
Research 10, uhad007. doi: 10.1093/hr/uhad007. 

Xin, T., Zhang, Z., Li, S., Zhang, S., Li, Q., Zhang, Z.-H., et al. (2019). Genetic Regulation of 
Ethylene Dosage for Cucumber Fruit Elongation. Plant Cell 31, 1063–1076. doi: 
10.1105/tpc.18.00957. 

Xing, Y., Cao, Y., Ma, Y., Wang, F., Xin, S., and Zhu, W. (2023). QTL mapping and 

transcriptomic analysis of fruit length in cucumber. Frontiers in Plant Science 14. 

Xu, X., Lu, L., Zhu, B., Xu, Q., Qi, X., and Chen, X. (2015). QTL mapping of cucumber fruit 

flesh thickness by SLAF-seq. Sci Rep 5, 15829. doi: 10.1038/srep15829. 

Yang, C., Li, H., Zhang, J., Luo, Z., Gong, P., Zhang, C., et al. (2011). A regulatory gene induces 
trichome formation and embryo lethality in tomato. Proceedings of the National Academy 
of Sciences 108, 11836–11841. doi: 10.1073/pnas.1100532108. 

Yang, S., Cai, Y., Liu, X., Dong, M., Zhang, Y., Chen, S., et al. (2018). A CsMYB6-CsTRY 

module regulates fruit trichome initiation in cucumber. J Exp Bot 69, 1887–1902. doi: 
10.1093/jxb/ery047. 

Yang, S. L., Pu, H., Liu, P. Y., and Walters, T. W. (1991). Preliminary studies on Cucumis 

sativus var. xishuangbannanesis. Cucurbit Genetics Cooperative, 10. 

Yashiro, K., Shimomura, K., Duong, T.-T., Shrestha, D. S., Sharma, S., Joshi, G. D., et al. 

(2017). Collaborative Exploration of Cucurbitaceae Vegetable Genetic Resources in 
Western Nepal, in 2016. 

Yu, J., Wu, S., Sun, H., Wang, X., Tang, X., Guo, S., et al. (2023). CuGenDBv2: an updated 

database for cucurbit genomics. Nucleic Acids Research 51, D1457–D1464. doi: 
10.1093/nar/gkac921. 

Yuan, X. J., Pan, J. S., Cai, R., Guan, Y., Liu, L. Z., Zhang, W. W., et al. (2008). Genetic 
mapping and QTL analysis of fruit and flower related traits in cucumber (Cucumis 
sativus L.) using recombinant inbred lines. Euphytica 164, 473–491. doi: 
10.1007/s10681-008-9722-5. 

Zhang, H., Wang, Y., Tan, J., and Weng, Y. (2022). Functional copy number variation of 

CsSHINE1 is associated with fruit skin netting intensity in cucumber, Cucumis sativus. 
Theor Appl Genet 135, 2101–2119. doi: 10.1007/s00122-022-04100-4. 

Zhang, J., Feng, S., Yuan, J., Wang, C., Lu, T., Wang, H., et al. (2021a). The Formation of Fruit 

Quality in Cucumis sativus L. Frontiers in Plant Science 12. doi: 
10.3389/fpls.2021.729448. 

35 

 
Zhang, K., Yao, D., Chen, Y., Wen, H., Pan, J., Xiao, T., et al. (2023). Mapping and 

identification of CsSF4, a gene encoding a UDP-N-acetyl glucosamine-peptide N-
acetylglucosaminyltransferase required for fruit elongation in cucumber (Cucumis sativus 
L.). Theor Appl Genet 136, 54. doi: 10.1007/s00122-023-04246-9. 

Zhang, S., Liu, S., Miao, H., Wang, M., Liu, P., Wehner, T. C., et al. (2016). Molecular Mapping 
and Candidate Gene Analysis for Numerous Spines on the Fruit of Cucumber. Journal of 
Heredity 107, 471–477. doi: 10.1093/jhered/esw028. 

Zhang, Y., Shen, J., Bartholomew, E. S., Dong, M., Chen, S., Yin, S., et al. (2021b). TINY 

BRANCHED HAIR functions in multicellular trichome development through an ethylene 
pathway in Cucumis sativus L. The Plant Journal 106, 753–765. doi: 10.1111/tpj.15198. 

Zhang, Z., Wang, B., Wang, S., Lin, T., Yang, L., Zhao, Z., et al. (2020). Genome-wide Target 
Mapping Shows Histone Deacetylase Complex1 Regulates Cell Proliferation in 
Cucumber Fruit. Plant Physiology 182, 167–184. doi: 10.1104/pp.19.00532. 

Zhao, J., Jiang, L., Che, G., Pan, Y., Li, Y., Hou, Y., et al. (2019). A Functional Allele of 
CsFUL1 Regulates Fruit Length through Repressing CsSUP and Inhibiting Auxin 
Transport in Cucumber. Plant Cell 31, 1289–1307. doi: 10.1105/tpc.18.00905. 

Zhao, J.-L., Pan, J.-S., Guan, Y., Zhang, W.-W., Bie, B.-B., Wang, Y.-L., et al. (2015). Micro-

trichome as a class I homeodomain-leucine zipper gene regulates multicellular trichome 
development in Cucumis sativus. J Integr Plant Biol 57, 925–935. doi: 
10.1111/jipb.12345. 

Zhou, G., Chen, C., Liu, X., Yang, K., Wang, C., Lu, X., et al. (2022). The Formation of Hollow 

Trait in Cucumber (Cucumis sativus L.) Fruit Is Controlled by CsALMT2. International 
Journal of Molecular Sciences 23, 6173. doi: 10.3390/ijms23116173. 

Zhou, Q., Miao, H., Li, S., Zhang, S., Wang, Y., Weng, Y., et al. (2015a). A Sequencing-Based 

Linkage Map of Cucumber. Molecular Plant 8, 961–963. doi: 
10.1016/j.molp.2015.03.008. 

Zhou, Q., Wang, S., Hu, B., Chen, H., Zhang, Z., and Huang, S. (2015b). An ACCUMULATION 

AND REPLICATION OF CHLOROPLASTS 5 gene mutation confers light green peel in 
cucumber. Journal of Integrative Plant Biology 57, 936–942. doi: 10.1111/jipb.12355. 

36 

 
 
 
 
CHAPTER 2  

Identification of QTL Associated with Resistance to Phytophthora Fruit Rot in Cucumber 
(Cucumis sativus L.) 

Identification of QTL associated with resistance to Phytophthora fruit rot in cucumber (Cucumis 
sativus L.) 
Ying-Chen Lin, Ben N. Mansfeld, Xuemei Tang, Marivi Colle, Feifan Chen, Yiqun Weng, 
Zhangjun Fei, Rebecca Grumet 

This work is accepted for publication in ‘Frontiers in Plant Science’ 

All of the work presented and writing was performed by YC Lin with the exception of the age-
related resistance QTL-seq and RNA-seq performed by B Mansfeld and M Colle. F Chen and Y 
Weng performed line development for the cucumber core population. X Tang and Z Fei 
performed bioinformatic analysis and SNP calling for the re-sequenced core population lines used 
for the GWAS and XP-GWAS analyses. 

37 

 
 
 
 
 
 
 
 
  
Abstract 

Phytophthora fruit rot (PFR) caused by the soilborne oomycete pathogen, Phytophthora 

capsici, can cause severe yield loss in cucumber. With no resistant variety available, genetic 

resources are needed to develop resistant varieties. The goal of this work was to identify 

quantitative trait loci (QTL) associated with resistance to PFR using multiple genomic approaches 

and populations. Two types of resistances have been identified:  age-related resistance (ARR) and 

young fruit resistance.  ARR occurs at 12-16 days post pollination (dpp), coinciding with the end 

of exponential fruit growth. A major QTL for ARR was discovered on chromosome 3 and a 

candidate gene identified based on comparative transcriptomic analysis.  Young fruit resistance, 

which is observed during the state of rapid fruit growth prior to commercial harvest, is a 

quantitative trait for which multiple QTL were identified.  The largest effect QTL, qPFR5.1, 

located on chromosome 5 was fine mapped to a 1-Mb region.  Genome-wide association studies 

(GWAS) and extreme-phenotype genome-wide association study (XP-GWAS) for young fruit 

resistance were also performed on a cucumber core collection representing > 96% of the genetic 

diversity of the USDA cucumber germplasm.  Several SNPs overlapped with the QTL identified 

from QTL-seq analysis on biparental populations.  In addition, novel SNPs associated with the 

resistance were identified from the germplasm. The resistant alleles were found mostly in 

accessions from India and South Asia, the center of diversity for cucumber. The results from this 

work can be applied to future disease resistance studies and marker-assisted selection in breeding 

programs. 

Introduction 

Cucumber production in the United States primarily serves two markets: slicing (fresh) and 

pickling (processing).  Pickling cucumber production using primarily the once-over machine 

38 

 
harvesting system in the Midwestern U.S. is severely impacted by Phytophthora fruit rot caused 

by the soil-borne oomycete, Phytophthora capsici (Savory et al., 2011; Hausbeck and Lamour, 

2004).  The pathogen primarily infects cucumber fruits, especially young, rapidly growing fruits, 

while other plant parts such as leaves and vines remain intact (Gevens et al., 2006).  The 

cucumber plants grown for processing are planted on bare ground in high density to facilitate 

machine harvest.  The fruits, which are beneath the vegetative canopy, are in direct contact with 

the infested soil and the high density of foliage creates a moist environment ideal for the growth 

and spread of P. capsici (Ando and Grumet, 2006; Ando et al., 2007).  The pathogen releases 

flagellate zoospores from sporangia that are mobile in water and can easily spread within or 

between fields through irrigation or rainwater (Granke et al., 2012).  The symptoms start with 

water-soaked lesions and tissue collapse, followed by the growth of white mycelia and sporangia 

on the fruit surfaces (Quesada-Ocampo et al., 2023).  Currently, growers rely on a combination of 

exclusion, cultural practices for management, and chemical control strategies to reduce outbreaks 

of P. capsici (Sanogo et al., 2022; Quesada-Ocampo et al., 2023).  However, P. capsici can 

readily develop resistance to chemical controls and several commonly used fungicides such as 

metalaxyl and mefenoxam have been shown to be ineffective against the pathogen (Kousik et al., 

2017; Quesada-Ocampo et al., 2023).  Growing resistant varieties could reduce the use of 

fungicides, which should lead to lower cost and less potential environmental or health hazards.  

However, the complex nature and genetic variation in both host and pathogen hamper the 

development of resistant commercial cultivars. In cucumber, sources of resistance were 

discovered (Gevens et al., 2006; Colle et al., 2014), but are yet to be introgressed into commercial 

varieties.  

The effort of searching for resistance to P. capsici initially led to the discovery of age-

39 

 
related resistance (ARR) (Gevens et al., 2006; Ando et al., 2009). The fruits from accessions 

expressing ARR (ARR+) are susceptible at early fruit development stages, then gradually become 

resistant as fruits develop (Ando et al., 2015).  The transition begins at ~12 days post pollination 

(dpp), toward the end of exponential fruit growth, and was demonstrated to be associated with the 

fruit peel (Ando et al., 2012, 2015).  Preformed biochemical defenses and metabolites that are 

developmentally regulated were found to be associated with ARR, these include enzymes 

producing defense related compounds such as reactive oxygen species and terpenoid glycosides 

(Mansfeld, et al., 2017).  Resistant-aged fruit also appear to be uniquely able to sense the presence 

of P. capsici zoospores, as evident by a spike in defense response genes as early as 2 hours post 

inoculation. This corresponds with observable death of zoospores early as in 4 hour-post-

inoculation in fruits that exhibit ARR (Mansfeld et al., 2020).  

While ARR is a largely effective form of resistance, cucumber fruits are usually harvested 

and consumed during mid- to late- exponential growth (approximately 8-12 dpp), prior to the 

transition to ARR.  Hence, cucumber growers will still suffer yield loss due to P. capsici even if 

the cultivars exhibit ARR.  Therefore, resistance expressed before the harvest age is desired to 

alleviate the yield loss caused by the pathogen.  To search for young fruit resistance, Colle et al. 

(2014) surveyed the U.S. cucumber plant introduction (PI) by testing young cucumber fruits (~5-7 

dpp).  Three accessions with low disease scores were potential sources of young fruit resistance.  

One of them, PI 109483 from Türkiye, exhibited stable resistance in the following generations of 

selfing.  The resulting S6 progeny was released as a breeding line MSU 109483-53 (Grumet and 

Colle, 2017).  The breeding line was later used for doubled haploid (DH) production, and the DH 

line ‘A4-3’, which shows delayed and reduced symptoms and slower rate of pathogen growth, 

was used to study young fruit resistance (Zhang et al., 2021).  

40 

 
The objectives of this work were to Identify genetic loci associated with both young fruit 

resistance and ARR and to develop molecular markers for future breeding efforts.  Multiple 

genetic and genomic approaches, including bulk segregant analyses, fine mapping, transcriptome 

analyses and genome wide association studies identified several loci in association with the 

resistance traits.  Association analyses identified several SNPs that overlapped with QTL 

identified from QTL-seq analysis as well as novel SNPs and potential sources of resistance.  

Materials and Methods 

Plant materials 

Biparental mapping.  The source of young fruit resistance was MSU109483-53 (Grumet and 

Colle, 2017), a breeding line obtained through a series of pure line selections from PI 109483, a 

landrace collected from Türkiye.  Seed from MSU109483-53 was used for doubled haploid (DH) 

production via in vivo-induced parthenogenic embryo culture generously performed by Rijk 

Zwaan (De Lier, Netherlands).  The DH line ‘A4-3’ was crossed with the susceptible parent, 

‘Gy14’, an American type pickling cucumber, which has been broadly used in research and 

breeding programs and for which a high-quality reference genome is available 

(http://cucurbitgenomics.org/v2/).  ‘A4-3’ also was crossed with an American fresh market 

cucumber, ‘Poinsett 76’, which exhibits ARR (Mansfeld et al., 2020), to provide a second 

population for QTL verification.  To map ARR, ‘Gy14’ (ARR-) was crossed with ‘Poinsett 76’ 

(ARR+) and resultant F1 seed was sent for DH production (Rijk Zwaan, De Lier, Netherlands).   

Cucumber core collection. The cucumber core collection was selected based on genotyping-by-

sequencing (GBS) data of United States National Plant Germplasm System (NPGS) collection 

(Wang et al., 2018).  To reduce heterozygosity and heterogeneity within the accessions, 

individuals from each accession were self-pollinated for 2 or 3 generations. The self-pollinated 

41 

 
core collection lines were re-sequenced at 30-40× coverage and used to call SNPs as described by 

Yu et al. (2023) (http://cucurbitgenomics.org/v2/).  The core collection lines were grown in the 

field from 2019-2022 with three plants per accession.  The accessions that were tested are listed in 

Supplementary Table 2.1.  The number of accessions planted each year varied depending on 

seed availability; each accession has 1-4 years of phenotypic data. 

Growth conditions 

Seeds for all experiments were sown in the greenhouse or growth room. Seedlings were 

transplanted to the greenhouse or field at the two true-leaf stage.  No fungicide was applied after 

the onset of flowering in either the greenhouse or field to ensure that fungicide residue was not 

present on the fruit surface to interfere with phenotyping. 

Greenhouse. Seedlings were transplanted to 1.5-gallon pots with Suremix Perlite soil medium 

and grown in the Michigan State University Plant Science Greenhouse Complex. The plants were 

fertigated twice a day (44 ppm nitrogen of Peters Professional 20-20-20 General Purpose; Scotts, 

Marysville, OH).  LED lights were used to provide 16-hour photoperiod.  Pest and disease 

management was based on general practice in the greenhouse using a combination of chemical 

and biological controls.  For young fruit experiments, bumble bees (Koppert Biological Systems, 

Inc., Howell, MI) were introduced at week 4-5 for pollination.  For ARR experiments, flowers 

were hand-pollinated; a single fruit was set per plant to prevent developmental effects of 

competition among fruits.  All ARR experiments were grown in the greenhouse.   

Field. Plants were grown at the Michigan State University Horticulture Teaching and Research 

Center (HTRC).  Prior to transplanting, 300 lbs/acre of 19-19-19 fertilizer were applied to the 

field, and irrigation provided as needed throughout the season.  To minimize contamination with 

other pathogens and avoid injury resulting from washing soil from the fruit, plants were grown on 

42 

 
raised black plastic mulch and trellised using T-posts and trellis netting. The space between plants 

was 0.45-0.6 m, depending on the field design each year.  Pollination was facilitated by 

honeybees. 

Screening for response to P. capsici 

Fruit harvesting and handling. To provide uniform, high-inoculum pressure and optimal 

environmental conditions for disease development, fruits grown in both the greenhouse and field 

were harvested at the desired stage of development and brought into the laboratory for disease 

screening.  For young fruit experiments, fruit were harvested at early exponential growth stage, 5-

7 dpp (~7-10 cm long).  Harvests were performed 2-3 times a week to provide 30-60 

fruit/accession for the core collection, or 10-30 fruit/plant for biparental segregating populations.  

For ARR experiments, fruit were harvested at 16-18 dpp.  Fruit from the field were rinsed with 

distilled water to remove soil and debris, sanitized by soaking in 1% bleach for one minute, and 

rinsed with distilled water thoroughly to remove bleach residue. Greenhouse fruit were rinsed 

with distilled water. Clean, dry fruits were placed in covered plastic trays lined with wet paper 

towel on the sides to maintain high humidity for pathogen growth as described by Gevens et al. 

(2006).  Trays were incubated at 25-26 °C under constant light. 

Pathogen inoculation and phenotyping. The P. capsici isolates, Bartley’s 1, OP97, or NY-0644-

RFP (Dunn et al., 2013), were cultured on V8 agar media as described in Gevens et al. (2006).  

After seven days, the plate was flooded with 6-7 mL sterile distilled water to stimulate zoospore 

production.  The concentration of resuspended zoospore was measured using a CountessTM 

automated cell counter (Thermo Fisher Scientific, Waltham, MA).  For young fruit experiments, 

zoospore suspensions of Bartley’s 1 were diluted to 1×104 zoospore/mL.  Two 30 μL droplets 

were applied to the surface of each fruit as described by Colle et al. (2014).  Inoculated cucumber 

43 

 
fruits were photographed and scored at 5 days post inoculation (dpi) based on the disease rating 

scale shown in Figure 2.1.  Ratings of 1-3 indicate mild symptoms limited the region of 

inoculation, 4-6 moderate to extensive water soaking, and 7-9 visible hyphal growth and 

sporulation.  Symptoms were scored at both inoculated sites on each fruit; the score for an 

individual fruit was the mean of the two sites. The disease rating for a plant or line was the 

average of all fruit over all harvests for a given experiment. For the ARR experiments, OP97 or 

NY-0644-RFP zoospore suspensions were diluted to 1×105 zoospores/mL.  Disease symptoms 

were monitored daily for 7-10 days and rated using either a 1-9 disease rating scale as above (F2 

experiments) or a 0-5 point disease score (0 – no symptoms, 5 – severe sporulation) (DH 

experiments).  Fruits from the DH population were inoculated with 12 equally spaced 30 μL 

droplets.  At 7 dpi, each fruit was assigned the rank of the most susceptible inoculation site. 
Highly susceptible
Moderately susceptible

Resistant

1

2

3

4

5

6

7

8

9

Mild symptoms in the regions
Mild symptoms in the
of inoculation
regions of inoculation

Moderate to extensive water 
Moderate to extensive
soaking
water soaking

Hyphal growth and sporulation
Hyphal growth and
sporulation

Fig. 1. The 9-point disease scoring scale of P. capsici infection on cucumber young fruits. 

Figure 2.1. Illustration of the 9-point disease scoring scale of Phytophthora capsici infection 
on cucumber fruit. Ratings 1- 3: no or minor symptoms limited to inoculation sites; ratings 4-6: 
levels of water soaking and necrosis; ratings 7-9: different levels of hyphal growth and 
sporulation. 

QTL-seq analysis 

Young fruit resistance.  An F2 population (n=362) from the ‘Gy14’ × ‘A4-3’ was screened in 

2018 in the field.  Leaf tissue (~50mg) was collected from each seedling prior to transplanting to 

44 

 
 
 
the field, freeze-dried, and ground for DNA extraction.  DNA was extracted from the parental 

lines and each plant selected for the bulks using the Kingfisher DNA extraction robot and Mag-

Bind® Plant DNA DS 96 Kit (Omega Bio-tek, Norcross, GA) as described in Wang et al. (2018).  

The genomic DNA was quantified using PicoGreen (Invitrogen™, Waltham, MA).  Equal 

amounts of DNA of the selected individuals for each bulk (n=19) were mixed for sequencing.  

Sequencing was performed at the Research Technology Support Facility at Michigan State 

University.  Three libraries (resistant bulk, susceptible bulk, and ‘A4-3’) were prepared using 

Illumina TruSeq Nano DNA Library Preparation Kit and sequenced using paired-end sequencing 

of 150 bp on an Illumina HiSeq 4000 platform.  After removing low-quality reads and sequencing 

adaptors using Trimmomatic v. 0.33 (Bolger, 2014), the reads from each bulk were aligned to the  

‘Gy14’ reference genome v2.0 (http://cucurbitgenomics.org/v2/; Yu et al., 2023) using BWA-

MEM (v0.7.8) (Li, 2013) with default parameters.  Sequencing duplicates were marked using 

PicardTools (https://broadinstitute.github.io/picard/, v2.7.1) and the Genome Analysis Toolkit 

(GATK; v3.6) best practice pipeline was used for SNP calling (McKenna et al., 2010; DePristo et 

al., 2011; Van der Auwera et al., 2013).  QTL-seq analysis (Takagi et al., 2013) was performed 

using the R package, QTLseqr (Mansfeld and Grumet, 2018).  SNP filtering criteria in QTLseqr 

was set for “minimum depth ≥ 50, maximum depth ≤ 100, GQ ≥ 99, depth difference ≥ 20, 

minimum sample depth ≥ 20,” which left 587,178 SNPs for further analysis.  Delta SNP-index 

and G’-value (Magewene et al., 2012) were calculated using a sliding window size of 1Mb, where 

the 95% and 99% confidence intervals were calculated with 10,000 iteration in QTL-seq analysis, 

while filter method “deltaSNP” at the threshold of 0.1 in G’ analysis. 

ARR.  Progeny of 79 DH lines (5 plants/line) along with both parents and F1 seed were grown in 

the greenhouse in a replicated block design. In a separate experiment, 92 F2 plants of the ‘Gy14’ × 

45 

 
‘Poinsett 76’ cross were grown in the greenhouse; plants with the 15 highest and lowest disease 

scores were selected for the two bulks respectively.  DNA extraction was as described above.  

After quantitation, all libraries were pooled in equimolar amounts and loaded on one lane of an 

Illumina HiSeq 2500 High Output flow cell (v2) and sequenced in a 2 ×150bp paired end format.  

Reads were cleaned and adaptor sequences were removed using Trimmomatic v. 0.33 (Bolger et 

al., 2014).  QTL-seq analysis was performed as previously described with the following settings 

(F2/DH): refAlleleFreq = 0.1/0.1, minTotalDepth = 20/30, maxTotalDepth = 50/100, 

depthDifference = 10/30, minGQ = 30/30, minSampleDepth = 10/15.  A window size of 2 Mb 

was used for smoothing Δ(SNP-index) values. 

Verifying and narrowing the genomic regions 

Young fruit.  A second F2 population (‘Gy14’× ‘A4-3’; n=752) was used to verify the QTL 

regions and develop RIL and F3 populations.  Polymorphic SNPs flanking and within the QTL 

region were identified for KASP marker design and ordered from Sigma-Aldrich (St. Louis, MO).  

The 5 μL reaction mixture contained 2.5 μL of DNA (at 10 ng/μL), 2.5 μL of 2 × KASP Master 

Mix (LGC Biosearch Technologies, Hoddesdon, UK; 3CR Bioscience, Essex, UK), and 0.07 μL 

of primer mix.  The cycling conditions were as follows:  94°C for 15 min followed by 10 

touchdown cycles at 94°C for 20 s and 65°C for 60 s (decrease 0.6°C per cycle), then 38 

amplification cycles of 94°C for 20 s and 55°C for 60 s, finally with 37°C for 10 s.  

Thermocycling and fluorescence readings were performed using a FX384 Real-Time thermal 

cycler (BioRad, Hercules, CA), where allele calls were determined using the CFX manager 

software (v.3.1).  To narrow the genomic region, homozygous recombinant individuals (i.e., 

homozygous for the ‘Gy14’ allele at one and homozygous for the ‘A4-3’ allele at the other end) 

were selected and self-pollinated for 4-5 generations to develop a RIL population.  Individuals 

46 

 
that were partially heterozygous (heterozygous at one end, and homozygous for the ‘Gy14’ or 

‘A4-3’ allele at the other end) were selfed to produce F3 families.  KASP markers were designed 

at approximately every 0.5 Mb within the QTL region.  The sequences of KASP markers and 

targeted SNP locations are in Supplementary Table 2.2.  RIL families were grown in the field in 

2020 with 30-60 fruits tested per line.  In 2022, 99 homozygous recombinant individuals from 33 

F3 families were grown in the greenhouse.  Cuttings were also taken from each plant and 

transplanted to the field.  Fruit were harvested from both sets of plants and disease screening was 

performed as described above. 

ARR.  To verify the ARR QTL, KASP markers flanking the QTL were used to genotype 768 F2 

seedlings of ‘Gy14’ × ‘Poinsett 76’ as described above.  Individuals homozygous for ‘Gy14’ or 

‘Poinsett76’ alleles within the QTL region were self-pollinated to produce F4 lines.  Plants from 

14 F4 lines were grown in the greenhouse in a randomized complete block trial (5 plants/line).  

Phenotyping was performed as described above. 

RNA-seq analysis 

Sample collection and RNA extraction.  Flowers of plants from the ARR parental lines 

(‘Poinsett 76’, ‘Gy14’) were hand pollinated so that 8 and 16 dpp fruit were harvested on the 

same day.  Three fruit (biological replicates) were collected for each age and genotype. 

Uninoculated fruit peels were collected from 8 and 16 dpp fruit using a vegetable peeler and 

immediately frozen in liquid nitrogen. Samples were ground using a mortar and pestle in liquid 

nitrogen. RNA extraction was performed using the MagMAX Plant RNA Isolation Kit protocol 

(Thermo Fisher, Waltham, MA) as described in Mansfeld et al. (2020). Assessment of RNA 

concentration and quality was performed as described in Rett-Cadman et al. (2019).  All samples 

had a minimum RNA quality score of 8. 

47 

 
TruSeq Library preparation and sequencing.  Libraries were prepared at Michigan State 

University’s Research Technology Support Facility, using the Illumina TruSeq Stranded mRNA 

Library Preparation Kit on a Sciclone G3 robot following manufacturer’s recommendations.  An 

additional cleanup with 0.8 × AmpureXP magnetic beads was performed after completion of 

library preparation.  Quality control and quantification of completed libraries were performed 

using a combination of Qubit dsDNA HS and Advanced Analytical Fragment Analyzer High 

Sensitivity DNA assays.  The libraries were divided into two pools of 15 libraries each.  Pools 

were quantified using the Kapa Biosystems Illumina Library Quantification qPCR kit.  Each pool 

was loaded onto one lane of an Illumina HiSeq 4000 flow cell and sequencing was performed in a 

1 × 50 bp single read format using HiSeq 4000 SBS reagents.  Base calling was done by Illumina 

Real Time Analysis (RTA) v2.7.7 and output of RTA was demultiplexed and converted to FastQ 

format with Illumina Bcl2fastq v2.19.1. 

Differential expression analysis. Reads were cleaned, and adaptor sequences were removed 

using Trimmomatic v. 0.34 (Bolger et al., 2014) with the following settings: LEADING:3 

TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:35. Quality control was performed using 

FastQC (http://www.bioinformatics.bbsrc.ac.uk/projects/fastqc). A cucumber transcriptome fasta 

file was made from the ‘Chinese Long’ (v2) (Huang et al., 2009; Li et al., 2011) genome using the 

gffread function from the cufflinks software package (Trapnell et al., 2010) and high-quality reads 

were then quasi-mapped to the transcriptome using Salmon v. 0.9.1 (Patro et al., 2017) with 

default settings. Read quantification data was imported into R using the tximport R package 

(Soneson et al., 2015) and differential expression analysis was performed using DEseq2 (Love et 

al., 2014) with log-fold-change-shrinkage.  Age and genotype were combined into a single factor 

for differential expression analysis and contrasts between the four conditions (‘Poinsett 76’ 8 dpp, 

48 

 
‘Poinsett 76’ 16 dpp, ‘Gy14’ 8 dpp, ‘Gy14’ 16 dpp) were performed.  Differentially expressed 

genes were called significant using an adjusted p-value (Benjamini-Hochberg adjustment; false 

discovery rate) of less than 5% and an expression change of greater than two-fold was used to 

define biological significance. 

Association analysis 

GWAS.  The SNP data of the core collection was downloaded from CucGenDBv2 (Yu et al., 

2023).  SNPs were filtered using BCFtools (Danecek et al., 2021) and GATK (Van der Auwera et 

al., 2013) with the following criteria: bi-allelic, GQ scores >20, maximum read depth within one 

standard deviation of the mean read depth, minor allele frequency > 0.1, missing rate <20%, 

resulting in 1,168,270 SNPs for association analysis.  Marker-trait association analyses were 

performed using best linear unbiased estimates (BLUEs).  BLUEs were calculated using the R 

package lme4.  Association analysis was performed using GAPIT 3.0 (Genome Association and 

Prediction Integrated Tool) with MLM, FarmCPU, BLINK, and MLMM models implemented 

within the software (Wang and Zhang, 2021).  The significance threshold was calculated based on 

Bonferroni correction (p-value/N, N= number of SNPs used in the analysis), where the thresholds 

of adjusted p values of 0.05 and 0.01 corresponded to -log10(p) values of 7.368 and 8.06, 

respectively. 

XP-GWAS.  The 29 most resistant and 29 most susceptible accessions were selected based on 

2019-2021 phenotypic data and grown in the field in a randomized complete block design with 

three blocks in 2022. A random bulk of 29 accessions was selected from the full core population 

(Supplementary Table 2.1).  XP-GWAS analysis was performed as described in Yang et al. 

(2015).  In brief, reference and alternative allele depths at each SNP site were extracted from the 

core resequencing data and calculated for each bulk. The input data was then computed using the 

49 

 
R package, XP-GWAS, with the depth filter set at 500, which ended with 3,444,143 SNPs for the 

analysis.  The 5% false discovery rate (FDR) and the threshold of p = 0.05 with Bonferroni 

correction were calculated to detect significant SNPs. 

Results 

Identification of QTL for young fruit resistance 

QTL-seq analysis 

Screening for young fruit resistance to P. capsici was performed on an F2 population 

(n=362) derived from the cross between the susceptible pickling cucumber breeding line, ‘Gy14’, 

and the doubled haploid line, ‘A4-3’, which shows reduced and delayed symptom development in 

response to inoculation (Figure 2.2A).  Three harvests were performed with 10-20 young fruits 

(5-7 dpp; ~7-10 cm long) sampled from each F2 plant.  The normally distributed disease scores 

suggested that young fruit resistance is a quantitative trait controlled by multiple loci (Figure 

2.2B).  To verify phenotyping of plants to be selected for the resistant and susceptible bulk 

populations, the 30 highest and lowest scoring individuals from the first three harvests were 

harvested an additional time.  Based on the four harvests, 19 plants which had consistent 

phenotypes were selected for the resistant and susceptible bulks, respectively (Figure 2.2C).   

Sequencing of the bulk populations generated ~77 million reads for each bulk with > 60´ 

coverage of the cucumber genome.  After processing and filtering, 558,625 SNPs were available 

for QTL-seq analysis, which identified QTL on chromosomes 1, 5, and 6 (Figure 2.3 and Table 

2.1).  The most significant QTL was located on chromosome 5. 

QTL validation and narrowing the genomic region of qPFR5.1 

To validate the regions identified by QTL-seq, a second F2 population (n=752) was genotyped 

using KASP markers flanking the QTL.  Individuals homozygous for either susceptible 

50 

 
(A)

(B)

(C)

Figure 2.2. Screening of young fruit from the F2 population of ‘Gy14’ ´ ‘A4-3’ for response 
to inoculation with P. capsici.  Fruits were harvested at 5-7 days post pollination (dpp).  (A) 
Example of disease screening; fruits were photographed at 5 days post inoculation (dpi).  Red 
box, A4-3; Dashed box, ‘Gy14’.  (B) Disease score distribution of the F2 population.  Scores are 
the average of 10-20 fruit from each F2 plant; fruit were scored at 5 dpi.  (C) F2 individuals 
selected for the resistant and susceptible bulks.  Disease scores are the average of all four 
harvests. 

‘Gy14’ (S) or resistant ‘A4-3’ I alleles on chromosome 5 and/or 6 were selected, providing four 

allelic combinations (chr5-chr6: Gy-Gy, Gy-A4-3, A4-3-Gy, and A4-3-A4-3).  Presence of the 

‘A4-3’ allele on chromosome 5 was associated with resistance with a frequency of 0.89 in the 

most resistant plants (disease score < 4) progressively dropping to 0.06 in the most susceptible 

plants (disease score > 7) (Figure 2.4A).  Individuals with ‘A4-3’ alleles at chromosome 5 

showed significantly lower disease score compared to those with ‘Gy14’ allele (Figure 2.4B). No 

significant allelic effect was observed for the QTL on chromosome 6 (Figure 2.4B).  The allelic 

effect of the QTL on chromosome 1 was not verified.  

The QTL on chromosome 5, named qPFR5.1 (Phytophthora fruit rot 5.1), was tested in a 

51 

 
  
 
 
 
Figure 2.3. QTL associated with young fruit resistance as identified by QTL-seq.  Δ(SNP-
index) and G’ were calculated with a window size of 1 Mb. Horizontal lines in Δ(SNP-index) 
represent confidence thresholds of 95% (red) and 99% (blue); for G’ the threshold for false 
discovery is 0.01 (blue line).  Allele frequencies for the resistant and susceptible bulks are 
provided in Supplementary Figure 2.1. 

52 

 
 
 
 
Table 2.1. QTL identified from QTL-seq in the ‘Gy14’ x ‘A4-3’ F2 population. 

QTL-seq 

G’ 

Chromosome 

Location (Mb)1 

Length 

Location (Mb) 

Length (Mb) 

1 

5 

5 

6 

5.35-6.27 

26.87-27.71 

29.24-30.19 

26.92-28.95 

0.92 

0.84 

0.95 

2.03 

5.53-6.00 

23.40-30.94 

- 

0.47 

7.53 

- 

27.04-28.38 

1.34 

1Genomic locations are according to Gy14 v. 2.1 (CuGenDB v.2; 
http://cucurbitgenomics.org/v2/) 

second genetic background, ‘Poinsett 76’, a North American fresh market cucumber.  F2 

individuals from ‘Poinsett 76’ × ‘A4-3’ (n=768) were genotyped with flanking markers for 

qPFR5.1, and individuals homozygous for either ‘Poinsett 76’ or ‘A4-3’ in that region were self-

pollinated.  Plants from the resulting 25 F3 families were grown in the greenhouse and field and 

young fruit were harvested for inoculation.  Consistent with the RIL population, the ‘A4-3’ allele 

showed strong association with resistance in the greenhouse and field, respectively), further 

confirming the effect of qPFR5.1 (Figure 2.4C). 

qPFR5.1 as identified by QTL-seq and G’ spanned ~7Mb on chromosome 5.  To facilitate 

fine mapping, we tested a recombinant inbred line (RIL) population and F3 families selected for 

recombination within the QTL region.  F2 plants that were homozygous recombinant at qPFR5.1 

(i.e., homozygous for the ‘Gy14’ allele at one end and homozygous for the ‘A4-3’ allele at the 

other end) were self-pollinated to the F4 and F5 generations; nine resulting RILs were grown in the 

field in 2020 with 30-60 fruits tested per line.  Each plant was also genotyped with KASP markers 

53 

 
 
 
at approximately 0.5 Mb intervals within the QTL region (Figure 2.5A).  Based on genotypic and 

phenotypic data, the region was narrowed to 3.22 Mb between markers M2 and M5 (25.17-28.39 

Mb). 

To further refine the QTL, an additional 768 F2 plants were genotyped and 178 partially 

homozygous F2 individuals (i.e., heterozygous at one end and homozygous for ‘A4-3’ or ‘Gy14’ 

alleles at the other end) were self-pollinated.  Twelve F3 families were selected for each of the 

four genotypic combinations (heterozygous-‘Gy14’, ‘Gy14’-heterozygous, heterozygous-‘A4-3’, 

and ‘A4-3’-heterozygous) and 16 individuals per family were genotyped.  Of those, 99 

homozygous recombinant individuals from 33 families were transplanted to the greenhouse.  

Cuttings were also taken from each plant and transplanted to the field for testing in a second 

environment.   Additional KASP markers were designed between M2 and M5 (Figure 2.5B).  

The QTL identified from F3 families was located between M53-M58 (26.09-27.13 Mb; 1.04 Mb) 

with a slight difference between the two seasons tested; M54-M58 (26.32-27.13 Mb) in the 

greenhouse, and M53-M41 (26.01-26.85 Mb) in the field. 

Association analysis of resistance to P. capsici 

GWAS of the core collection 

The re-sequenced cucumber core collection, consisting of 388 accessions along with 

several breeding lines with important agronomic traits, represents >96% of genetic diversity in the 

U.S. NPGS (Wang et al., 2018; Yu et al., 2023).  The collection was planted in the field and fruits 

were tested from 2019 to 2021.  The number of accessions grown each year varied depending on 

seed availability.  Phenotypic data was collected from 370 accessions with 1-3 years of disease 

scores per accession. The normally distributed disease score of the core collection further 

affirmed that young fruit resistance is a quantitative trait (Figure 2.6A); the correlation between 

54 

 
years ranged from 0.48-0.80.  

GWAS analysis was performed using BLUE values calculated from disease scores from 

2019-2021 with one single-locus model (MLM) and three multi-locus models (FarmCPU, BLINK 

and MLMM) (Figure 2.6B, Table 2.2).  A total of 11 SNPs were identified from the different 

models:  seven in FarmCPU, five in BLINK, one in MLMM, and five in MLM.  The phenotype 

variance explained (PVE) of the SNPs ranged from 0.38-24.49%.  Several significant SNPs were 

identified in at least two models.  S1_21117743 (A/G) was significant in BLINK and MLM 

models with PVE of 9.01% and 7.74%, respectively. S2_10226744 (A/G) was detected in 

FarmCPU, MLMM, and MLM models, and a closely located SNP 27.69 kb upstream was 

detected in the BLINK model.  This was the only significant SNP identified in the MLMM model 

with a PVE of 24.49%.  Another SNP, S3_37752706 (C/T), was detected in FarmCPU and 

BLINK models.  

Phenotypes were significantly different between accessions carrying homozygous 

reference vs. alternate alleles for all significant SNPs except S3_18480786 (Supplementary 

Figure 2.2).  Of the nine SNPs, five alternate alleles led to increased resistance (lower disease 

scores).  Of those, only two of the alternate alleles were present in ‘A4-3’ (SNP S1_21117743 and 

S3_37752706), suggesting that the other alleles identified by GWAS may provide additional 

sources of resistance.  When the alternate alleles associated with lower disease scores were rare in 

the core collection (< 10%, i.e., < 38 accessions), the majority of accessions (64%-81%) carrying 

the alternate allele originated from the India/South Asia region (e.g., S1_21117743, 

S5_23563699, and S6_29175300) (Table 2.3).  Conversely, four of the five SNPs associated with 

increased resistance (S1_21117743, S4_8024257, S5_23563699, and S6_29175300) were very 

uncommon in the East Asian accessions (0-3%).  For S2_10226744, where the rare alternate allele 

55 

 
was associated with increased susceptibility, 77% of the accessions were from East Asia.  When 

the alternate alleles occurred frequently in the germplasm (>50%) (e.g., S3_37752706 and 

S7_3391182), the origins were widely distributed across regions. 

(A)

(B)

(C)

Figure 2.4. Verification of the allelic effect of QTL for young fruit resistance identified by 
QTL-seq analysis. (A) Disease distribution at 5 days post inoculation for the selected F2 plants 
(‘Gy 14’ × ‘A4-3’) homozygous for the ‘Gy14’ or ‘DH4-3’ alleles (n=82) (bar graph), and the 
frequency of the resistant ‘A4-3’ allele on chromosome 5 (black line).  Each F2 value is the mean 
of 10-30 fruit/plant.  (B) Disease scores of F2 individuals of ‘Gy14’ (pickling cucumber) ´ ‘A4-3’ 
possessing either the ‘A4-3’ or ‘Gy14’ allele at qPFR5.1. (C) F3 families of ‘Poinsett 76’ (fresh 
market cucumber) ´ ‘A4-3’ possessing either the ‘A4-3’ or ‘Poinsett’ allele at qPFR5.1.  Each 
point is the mean of >20 fruits/family from the greenhouse and >50 fruits/family from the field. 

56 

 
 
 
 
 
A

B

qPFR5.1

Field

GH

Figure 2.5. Fine mapping of qPFR5.1 in (A) RIL population and (B) F3 families. Dark, white, 
grey bars represent ‘A4-3’, ‘Gy14’, and heterozygous alleles, respectively.  Letters on the right 
indicate phenotypes of each line/family: r, resistant (score < 4.5); s, susceptible (score > 6.0); I, 
intermediate (scores 4.5 – 6.0). Genomic locations shown in parentheses for each KASP marker 
(Mb) as per Gy14 v. 2.1 (CuGenDB v.2; http://cucurbitgenomics.org/v2/). 

57 

 
 
 
 
 
Table 2.2. Significant SNPs identified in multiple GWAS models (FarmCPU, Blink, MLMM, and MLM) for young fruit 
resistance in the cucumber core collection. 

SNP 

Chr 

S1_21117743 

S2_10199046 

S2_10226744 

S3_18480786 

S3_37752706 

S4_8024257 

S5_23563699 

S6_29086459 

S6_29175300 

S7_3391182 

S7_17739386 

1 

2 

2 

3 

3 

4 

5 

6 

6 

7 

7 

Position  

(bp) 1 

21,117,743 

10,199,046 

p-value / PVE (percent phenotypic variation explained) (%) 

FarmCPU 

BLINK 

MLMM 

MLM 

- 

- 

5.15E-13 / 9.01 

8.31E-10 / 3.63 

- 

- 

1.73E-8 / 7.74  

- 

10,226,744 

3.85E-5 / 3.60 

18,480,786 

1.06E-8 / 0.38 

- 

- 

37,752,706 

3.37E-11 / 1.53 

4.49E-10 / 2.09 

8,024,257 

- 

2.64E-9 / 2.25 

23,563,699 

1.73E-8 / 3.23 

29,086,459 

2.70E-8 / 2.03 

29,175,300 

3.31E-11 / 5.48 

3,391,182 

9.52E-9 / 0.79 

- 

- 

- 

- 

17,739,386 

- 

6.04E-9 / 1.35 

9.76E-10 / 24.49 

5.95E-9 / 1.44 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

- 

1Genomic locations are according to Gy14 v. 2.1 (CuGenDB v.2; http://cucurbitgenomics.org/v2/) 

58 

 
 
 
Table 2.3. Geographical origin of accessions carrying the alternate alleles for the significant 
SNPs for young fruit resistance to P. capsici as identified by GWAS. 

SNP 

b
t
c
e
f
f

E

a
c
i
r
f
A

S1_2111743 

S3_37752706 

S4_8024257 

S5_23563699 

S6_29175300 

↓ 

↓ 

↓ 

↓ 

↓ 

S3_18480786 

- 

S2_10226744 

S7_3391182 

S7_17739386 

↑ 

↑ 

↑ 

- 

7 

4 

- 

- 

1 

1 

5 

1 

e
p
o
r
u
E

- 

28 

24 

1 

2 

2 

2 

43 

17 

a
i
s
A

t
s
a
E

- 

82 

1 

1 

1 

23 

23 

79 

17 

Region of origina 

a
i
s
A

t
s
e

W

/
l
a
r
t
n
e
C

2 

13 

11 

1 

- 

2 

- 

30 

7 

a
i
s
A
h
t
u
o
S

/
a
i
d
n
I

9 

24 

18 

26 

27 

5 

2 

47 

15 

a
c
i
r
e
m
A

h
t
r
o
N

1 

13 

4 

3 

4 

1 

2 

37 

10 

y
e
k
r
u
T

2 

30 

32 

- 

- 

- 

- 

32 

17 

r
e
h
t
O

- 

- 

2 

- 

- 

- 

- 

2 

- 

l
a
t
o
T

14 

197 

96 

32 

34 

34 

30 

275 

84 

a The regions are as defined in Wang et al. (2018).   

b The effect of alternative alleles compared to reference alleles in disease score. ↓ - decreased 

disease score (more resistant); ↑ - increased disease score (more susceptible) 

59 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
(A)

(B)

Figure 2.6. (A) Disease score distribution for young fruit resistance to Phytophthora capsici 
from the cucumber core collection and BLUE distribution of combined data 2019-2021. The 
score for each accession is the mean of 30-50 fruits. (B) Manhattan plots and quantile–
quantile plots of the genome-wide association study analyses for young fruit resistance in the 
cucumber core population. The horizontal blue and red lines represent significance thresholds of 
Bonferroni -corrected P values of 0.05 and 0.01, respectively. The dotted vertical lines show the 
locations of SNPs that were significant in at least two models.  

60 

 
 
XP-GWAS of the extreme phenotype bulks 

Precise phenotyping is crucial to identify QTL associated with traits of interest, especially 

for quantitative traits composed of multiple small effect QTL. However, depending on the trait, 

phenotyping can be expensive, especially when screening a diversity panel with many lines. To 

determine reproducibility of QTL identified from GWAS and increase replication and accuracy of 

phenotyping, we used an XP-GWAS approach (Yang et al., 2015). The accessions with extreme 

resistant or susceptible phenotypes (29 accessions in each bulk) were retested for additional 

phenotyping in 2022.  

The disease score distributions of the resistant and susceptible bulks showed clear 

differences in multiple years (Figure 2.7A) and was reproduced in the replicated trial in 2022 

(Figure 2.7B), verifying accuracy of the bulk selection for XP-GWAS analysis.  Correlations for 

the selected resistant and susceptible bulks among 2019, 2021, 2022 were 0.755-0.912.  SNP data 

from the selected accessions were combined via in-silico bulking as described in methods.  XP-

GWAS analysis identified 165 significant SNPs (5% FDR threshold) distributed across the seven 

chromosomes.  The 39 significant SNPs based on the Bonferroni corrected p=0.05 threshold were 

located on chromosomes 1 and 5 (Figure 2.7C, Supplementary Table 2.3).  The XP-GWAS 

SNP identified on chromosome 5 overlapped with the QTL identified by QTL-seq. 

Identification of QTL for age-related resistance 

QTL-seq analysis   

Screening for ARR was performed on two populations:  an F2 population from ‘Gy14’ 

(ARR-) (ARR+) (Figure 2.8A); and DH lines produced from F1 seed of ‘Gy14’ ´ ‘Poinsett 76’ 

(Figure 2.8B).  Plants were grown in the greenhouse and a single, hand-pollinated fruit per plant 

was harvested at 16-18 dpp.  The 15 most resistant and susceptible F2 individuals were selected 

61 

 
for QTL-seq analysis (mean disease ratings of 1.32 and 7.88 for resistant and susceptible bulks, 

respectively; 1-9 scale).  Disease scores for fruit from the 79 DH lines (3-5 fruit/line) showed high 

within-line variability for lines showing intermediate susceptibility; fruit from the most resistant 

and susceptible lines responded consistently with a mean rating of 0 and 4.8, respectively (scale 0-

5) (Supplementary Figure 2.3A).  Seed from the 15 most resistant and susceptible DH lines 

were regrown and fruit were phenotyped in a second screen (Supplementary Figure 2.3B). The 

eight lines with the most consistent resistant and susceptible disease ratings in both screens were 

selected.  

DNA from the 15 most resistant and susceptible F2 individuals and the 8 most resistant 

and susceptible DH lines were pooled for QTL-seq bulk segregant analysis.  A total of 72,699 and 

92,607 filtered SNPs were used in the F2 and DH analysis, respectively.  Both analyses identified 

a major locus associated with resistance on chromosome 3, qPARR3.1 (Phytophthora ARR 3.1) 

located at 34.62-38.07 and 31.08-41.68Mb, respectively (Figure 2.8A, B).  To verify the QTL, 

KASP markers flanking the QTL on chromosome 3 were used to genotype 768 F2 seedlings and 

individuals homozygous for ‘Gy14’ or ‘Poinsett 76’ within the QTL region were self-pollinated to 

produce F4 lines.  Phenotyping of fruit from a replicated trial (5 plants/F4 family) verified a strong 

effect of the QTL (Figure 2.8D).  

Identification of genes of interest within the linked locus 

Transcriptome analysis of the parental lines at 8 and 16 dpp was used to identify genes of 

interest within the region of qPARR3.1.  ARR results from developmental changes that occur 

prior to inoculation, either as a result of production of preformed resistance mechanisms, or a 

change in capacity to rapidly respond to infection (Mansfeld et al., 2017, 2022).  Therefore, we 

sought to identify developmental changes in gene expression that are unique to cultigens that 

62 

 
become resistant vs. those that remain susceptible.  Genes were considered of interest if they 

showed differential expression with age in ‘Poinsett 76’ and were also differentially expressed in 

‘Poinsett 76’ vs. ‘Gy14’ at 16dpp.  Of the 1,240 annotated genes in this region (CL9930v2; 

CuGenDB), only four genes were uniquely upregulated in resistant ‘Poinsett 76’ (i.e., up in 

‘Poinsett 76’ fruit peels at 16 dpp vs. ‘Poinsett’ 76 at 8 dpp, and in ‘Poinsett 76’ at 16 dpp vs. 

‘Gy14’ at 16dpp).  Thirteen genes were uniquely downregulated in resistant peels (down in 

‘Poinsett 76’ fruit peels at 16 dpp vs. ‘Poinsett 76’ at 8 dpp, and in ‘Poinsett 76’ at 16 dpp vs. 

‘Gy14’ at 16dpp) (Figure 2.9A; Supplementary Table 2.4).  

(B)

(A)

(C)

Figure 2.7. Disease score distribution and Manhattan plot of the XP-GWAS analysis to 
identify SNPs associated with young fruit resistance. (A) Disease score distribution of the 
resistant and susceptible bulks in different years. (B) Disease score values of the resistant I, 
susceptible (S), and random bulks (**** indicates P<0.0001, Wilcoxon test). (C) Manhattan plot 
of the XP-GWAS analysis. The dashed line indicates the 5% FDR threshold; the solid line 
indicates significance threshold of Bonferroni-corrected P value of 0.05. 

63 

 
 
 
 
 
(A)

(B)

8dpp

16dpp

(C)

(D)

(E)

Figure 2.8. QTL-seq analysis for age-related resistance (ARR) to Phytophthora fruit rot 
using segregating populations derived from ‘Gy14’ (ARR-) ´ ‘Poinsett 76’ (ARR+).  (A) 
Top: Disease rating distributions for F2 individuals (n=95).  All fruit were harvested at 16-18 dpp. 
Disease rating on a 1-9 scale. Fruit were scored at 10 dpi. The 15 most resistant and susceptible 
individuals were selected for QTL-seq.  Bottom: QTL-seq analysis.  Δ(SNP-index) was calculated 
with a window size of 2Mb.  Horizontal lines represent confidence thresholds of 95% (red) and 
99% (blue). (B) Top: Disease rating distributions for doubled haploid lines (n=79, 3-5 fruit/line).  
Disease ranking on 0-5 scale. Fruit were scored at 7 dpi.  The 8 most resistant and susceptible 
lines were selected for QTL-seq.  Bottom: QTL-seq analysis.  Δ(SNP-index) was calculated with 
a window size of 2Mb.  Allele frequencies for the resistant and susceptible bulks are provided in 
Sup. Fig. 1.  (C) Phenotype of ‘Poinsett 76’ fruit at 8 and 16 dpp photographed at 7 days post 
inoculation (dpi).  (D) Mean disease ratings of F4 families (5 plants/F4 family) homozygous for 
the ‘Gy14’ or ‘Poinsett 76’ allele within the QTL on chromosome 3.  I Disease ratings for 
‘Poinsett 76’ and ‘A4-3’ fruit harvested at 8 and 16 dpp and scored at 7 dpi.  Each value is the 
mean of 8-14 fruit. 

64 

 
 
 
 
To further examine genes potentially contributing to ARR, we screened our previously 

published transcriptome data from developing ‘Vlaspik’ fruit (ARR+) compared to ‘Gy 14’ fruit 

(ARR-) (Mansfeld et al., 2017) to identify genes uniquely up- or downregulated in both ARR+ 

cultivars at 16 dpp.  Of the 17 genes identified above, one upregulated gene and three 

downregulated genes had similar expression patterns in ‘Poinsett 76’ and ‘Vlaspik’ (Figure 

2.9B).  The up-regulated gene, Csa3G872720 (CsGy3G041010 in Gy14 v.2.1), showed 

consistently significantly higher expression in resistant ‘Poinsett 76’ and ‘Vlaspik’ at 16 dpp.  

Expression in 16 dpp ‘Poinsett 76’ fruit was >2.5-fold higher compared to 8 dpp susceptible 

‘Poinsett 76’ fruit and susceptible ‘Gy 14’ 16 dpp fruit.  Similarly, expression in resistant 

‘Vlaspik’ fruit was greater than 4-fold higher when compared to susceptible ‘Gy14’ 16 dpp fruit, 

and 1.8-fold higher when compared to susceptible ‘Vlaspik’ 8 dpp fruit.  In contrast, while the 

downregulated genes had statistically lower values in resistant fruit in both genotype comparisons, 

the patterns observed were not obviously consistent with our model of ARR; i.e., distinctly 

different expression levels in 16 dpp resistant fruit compared to the susceptible samples.  

CsGy3G041010 located at 38,365,049-38,373,217 bp (Gy14 v. 2.1), is annotated to 

encode a putative RING-type E3 ubiquitin transferase, a U-box domain and WD40 repeat 

containing protein.  Sequence comparisons from 2 kb upstream until the end of its 3’UTR in ‘Gy 

14’ (ARR-), ‘Poinsett 76’ (ARR+) and ‘Vlaspik’ (ARR+) identified 15 SNPs and 5 INDELs 

(Supplementary Table 2.5) that differed between ‘Gy14’ and ‘Poinsett 76’.  Four of the variants 

were within 2kb upstream of the transcription start site; eight were in introns, one of which was 

close to a splice site; three were in the 3’UTR.  Six SNPs were in exons, five of which cause non-

synonymous amino acid changes.  ‘Vlaspik’, which is a commercial F1 hybrid, was heterozygous 

at all variants.  In contrast to ‘Poinsett 76’ which shows a strong drop in disease score for 16 dpp 

65 

 
vs. 8 dpp fruit, the disease rating of ‘A4-3’ remained essentially constant at 8 dpp vs. 16 dpp 

indicating the distinct nature of ARR and young fruit resistance in these genotypes (Figure 2.8E).  

The CsGy3G041010 allele in ‘A4-3’ matches that of ‘Gy14’. 

(A)

(B)

Figure 2.9.  Identification of genes uniquely expressed in resistant fruit.  (A) Heatmap of 
genes within the locus identified as uniquely differentially expressed in resistant-aged (16 dpp) 
‘Poinsett 76’ fruit. Row clustering was based on Euclidean distances. Heatmaps are scaled by 
row and indicate deviation relative to mean expression across all samples. Gene names are 
according to Chinese Long v.2 (CuGenDB; http://cucurbitgenomics.org/). (Read count data and 
corresponding Gy14v.2.1 gene names are provided in Supplemental Table 4.) (B) Genes 
uniquely up- or down-regulated in transcriptome comparisons with two ARR+ genotypes 
(‘Poinsett 76’ and ‘Vlaspik’).  Boxplots show the distribution and median (dark line) of 
normalized read counts from two experiments. ‘Gy 14’ is susceptible at both ages (8 and 16 
dpp); ‘Poinsett 76’ and ‘Vlaspik’ are susceptible at 8 dpp and become resistant at 16 dpp. Three 
biological replicates of each age-genotype combination were used in each experiment. 

Discussion 

Employing multiple strategies to detect QTL for resistance to Phytophthora fruit rot 

The types of populations most frequently used to identify QTL associated with traits of 

interest are segregating biparental populations and genetically variable natural populations.  In 

this work, both types of populations were used to search for QTL associated with young fruit 

66 

 
 
 
resistance to Phytophthora fruit rot in cucumber.  QTL-seq, a bulk segregant analysis (BSA) 

approach performed on progeny from biparental populations that express extreme phenotypes, 

provides a quick and specific way to detect QTL.  Furthermore, using different population 

structures such as F2, RIL, and DH can provide additional power for QTL detection.  QTL-seq 

was performed on both young fruit resistance and ARR.  The analyses discovered three QTL 

associated with the young fruit resistance and one for ARR.  For young fruit resistance, the QTL 

were located on chromosomes 1, 5, and 6; the strongest effect was from the QTL on chromosome 

5, qPFR5.1.  For ARR, which appears to have a major gene component, a QTL at the end of 

chromosome 3 was identified.  The ability to identify QTL and the size of the genomic region 

identified is limited to the genetic variation between the two parents and is influenced by 

population structure and size (Mackay and Caligari, 2000; Li and Xu, 2021).  The lengths of QTL 

from QTL-seq can be large, for example, qPFR5.1 was ~7 Mb and qPARR3.1 was ~10 Mb. To 

narrow the QTL for young fruit resistance, screening of additional RIL and F3 populations that 

were enriched for recombination within the region refined qPFR5.1 to ~1 Mb.  

The second approach, use of a diversity panel such as the cucumber core collection, 

provides the opportunity to identify additional QTL associated with the resistances in a 

germplasm with higher genotypic diversity.  The main drawbacks of using a diversity panel 

include higher expenses to phenotype and genotype the large population size, and the difficulty to 

detect rare alleles associated with the traits (Alqudah, et al., 2020; Uffelmann et al., 2021).  In 

addition to traditional GWAS, where the whole collection is phenotyped and genotyped, a second 

method of association analysis, extreme-phenotype genome-wide association study (XP-GWAS) 

can be used to reduce experimental costs by reducing the number of entries to be included (Yang 

et al., 2015).  XP-GWAS is typically used to reduce sequencing costs and has been applied to 

67 

 
several crops with different target traits.  Some recent examples include rice (Xiao et al., 2017), 

apple (Kumar et al., 2022), switchgrass (Poudel et al., 2021), sugar beet (Ries et al., 2016), and 

wheatgrass (Crain et al., 2023).  In this case, however, as sequence data was already available for 

the full collection (Yu et al., 2023), we were able to use XP-GWAS to reduce phenotyping costs 

associated with additional replications.  By focusing on the lines with extreme phenotypes, the 

rare alleles associated with resistance can be enriched within the bulk and are more likely to be 

detected (Yang et al., 2015; Zou et al., 2016).  Although performing XP-GWAS still requires 

known phenotypes of each line, once candidates for the extremes are identified, the subsequent 

phenotyping expense for replication can be reduced.  In this research a preliminary screening of 

the complete cucumber core collection was performed from 2019-2021. To verify the phenotypes, 

selected accessions were grown in the following year with additional replications to increase the 

accuracy of phenotyping.  

It should be noted that when performing association analysis, the same data sets analyzed 

using different programs can give somewhat different results due to the default assumptions 

written within each software.  As a result, peak SNPs may be offset by several Mb.  For example, 

the significant SNP on chromosome 3 detected in rMVP (Ying et al., 2021), another widely used 

R-based GWAS software, was about 3 Mb away from the significant SNP detected on 

chromosome 3 in GAPIT3.0 (34,803,363 vs. 37,752,706, respectively).  Therefore, replication 

and comparison among tools and models are recommended to avoid false positives (Chanock 

et al., 2007).  Similarly, different experiments with standard QTL or QTL-seq analyses can give 

somewhat different estimates of QTL location (e.g., powdery mildew and downy mildew 

resistance QTL (Wang et al., 2020).  In other cases, QTL may be somewhat complex, composed 

of more than one contributing factor as was observed for cucumber downy mildew (Berg et al., 

68 

 
2020), possibly contributing to different assessments of QTL location.  These observations can 

have implications for consideration of appropriate regions for introgression of disease resistance 

QTL.   

Using multiple approaches, several SNPs significantly associated with resistance to 

Phytophthora fruit rot were identified in closely-located positions, giving greater confidence to 

their contributions (Figure 2.10).  In our results, regions were identified by more than one 

approach on chromosomes 1, 2, 3, 5, 6, and 7.  The QTL peaks detected on chromosome 1 were 

located at 21.17 Mb by GWAS and 24.75 Mb by XP-GWAS.  The QTL on chromosomes 5 and 6 

were consistently identified by QTL-seq, GWAS and XP-GWAS methods (on chromosome 5 all 

were located within 7 Mb, and on chromosome 6 all were within 3 Mb).  On chromosome 3, the 

peak SNP was located at 37.75 Mb in GWAS and at 39.29 Mb in XP-GWAS.  Both fall within 

the QTL region previously identified from QTL-seq analysis for ARR including the candidate 

gene CsGy3G041010 located at 38.36 Mb.  In most cases, the QTL identified for Phytophthora 

fruit rot also coincided with previously identified QTL for other diseases. 

The Identification of multiple QTL for young fruit resistance Is consistent with 

quantitative traits consisting of multiple small effect QTL.  Studies identifying QTL for resistance 

to P. capsici in other species also have indicated polygenic architecture, including several 

examples in pepper (Capsicum annum) and squash (Cucurbita pepo and Cucurbita moschata) 

(Barchenger et al., 2018; Ramos et al., 2020; Vogel et al., 2021).  The signal on chromosome 5 

was stronger in QTL-seq and XP-GWAS compared to GWAS, possibly due to the use of the 

resistant line ‘A4-3’ in the bi-parental QTL-seq analysis, and the enrichment of rare alleles in the 

resistant bulk for XP-GWAS analysis.   

69 

 
**

*
*

* *

1

.

i

5
R
F
P
q
g
n
p
p
a
m
e
n
F

i

R
R
A
e
n
e
g
e
t
a
d
d
n
a
C
*
*

S
A
W
G
-
P
X

i

S
A
W
G
*

l

a
t
n
e
r
a
p
b

i

L
T
Q

t
o
r

t
i
u
r
f
a
r
o
h
t
h
p
o
t
y
h
P

*
**

*
*

**

Figure 2.10.  Chromosomal locations of QTL identified for Phythophthora fruit rot of 
cucumber in relation to prior QTL identified for resistances to other cucumber diseases.  
The indicated PFR QTL were identified from multiple analyses: red bar – biparental QTL-seq; 
blue bar – fine mapping of biparental populations; red asterisk GWAS of cucumber core 
collection; blue asterisk XP-GWAS; purple asterisk – candidate gene identified by RNAseq 
analyses.  Figure is adapted from Wang et al., 2020 Horticulture Research under Creative 
Commons license http://creativecommons.org/licenses/by/4.0/.  Abbreviations:  DM, downy 
mildew; PM, powdery mildew; ALS, angular leaf spot; Foc, fusarium wilt; GSB, gummy stem 
blight; MYSV, melon yellow spot virus; CYSDV, cucurbit yellow stunting disorder virus. 

70 

 
 
 
 
 
 
 
 
 
Although the SNPs identified from the association analyses in this study were screened for 

young fruit resistance, qPARR3.1 from QTL-seq analysis was also detected by GWAS and XP-

GWAS analyses. This may result from closely located genes within the qPARR3.1 region that 

confer young fruit resistance. It may also be possible that for some accessions the ARR-associated 

QTL may function earlier during fruit development and contribute to young fruit resistance.  A 

BLAST search (Swiss-Prot database) of the protein sequence for the candidate gene for ARR on 

chromosome 3 reveals that this gene is a homologue of the LIN gene found in Medicago 

truncatula (E = 0e0, Identity = 58.38%, Positives = 72.86%).  In M. truncatula LIN functions in 

early rhizobial symbiotic nodule formation and mutation of LIN leads to a suppression of nodule 

development (Kiss et al., 2009). LIN was shown to not be required for nodule organogenesis, 

however LIN expression was associated with rhizobial root infection (Kiss et al., 2009).  

Interestingly, while other rhizobial-symbiosis mutants were more resistant to infection, the lin-2 

mutant was shown to be extremely susceptible to the pathogen Phytophthora palmivora (Rey et 

al., 2015).  It has been hypothesized that the protein modulates defense responses by means of its 

U-box domain and the ubiquitination of target proteins (Kiss et al., 2009). 

In addition to the QTL identified by QTL-seq in biparental populations, novel SNPs were 

also discovered by association analyses, providing potential additional resources for future 

breeding efforts. Multiple alternative alleles of the significant SNPs leading to stronger resistance 

were found in accessions originating from India and South Asia, the primary and secondary 

centers of origins of cucumber (Lv et al., 2012; McCreight et al., 2013).  During the process of 

subsequent domestication and dissemination, cucumber germplasm diverged between East Asia 

vs. Eurasia and the West (Europe, Africa, North America) (Qi et al., 2013; Wang et al., 2018).  

This divergence, which is evident in fruit morphology, is also reflected in genetic composition, 

71 

 
showing differentiation of cucumber into three major phylogenetic clades: India/South Asia, East 

Asia, and the West (Qi et al., 2013; Wang et al., 2018; Grumet et al., 2021).  Consistent with the 

resistance-associated alleles identified in this study, prior screening indicated that the accessions 

that were resistant to Phytophthora fruit rot were mainly from India, especially North and Central 

India (Colle et al., 2014; Grumet et al., 2020).  Although there were also resistant accessions that 

originated from East Asia, alleles associated with higher susceptibility were more frequently 

traced to East Asian accessions.   

QTL hotspots for disease resistance 

Numerous studies have been conducted to identify QTL of various disease resistances in 

cucumber as summarized in a recent review by Wang et al. (2020).  Collectively these studies 

implied the presence of disease resistance gene/QTL hot spots on several chromosomes.  

Interestingly, most of the QTL and significant SNPs identified here for Phytophthora fruit rot also 

co-localized with the hot spots including resistances to downy mildew, powdery mildew, fusarium 

wilt, and gummy stem blight (Figure 2.10).  The aggregation of these QTL suggests that these 

genomic regions play an important role in disease resistance to fungal and oomycete pathogens.  

Oomycete and fungi are two evolutionary distinct groups; however, they share similar strategies 

in terms of infection, e.g., specialized infection structures such as appressoria, infection hyphae 

and haustoria, and secreting cell-wall-degrading enzymes to facilitate cell wall penetration 

(Latijnhouwers et al., 2003). Though the mechanisms of resistances to these pathogens remain 

unknown, similar or the same defense-associated genes might be activated in response to infection 

and result in clusters of disease resistance QTL. 

In addition to Phythophora the other major disease leading to severe loss of pickling 

cucumber production in midwestern United States is downy mildew caused by 

72 

 
Pseudoperonospora cubensis (Savory et al., 2011; Hausbeck and Lamour, 2004).  Toward the end 

of chromosome 5 there are two disease resistance QTL clusters including resistances to powdery 

mildew and downy mildew pathogens.  qPFR5.1 for young fruit resistance identified in this work 

is located in one of these clusters and is adjacent to two downy mildew resistance QTL, dm5.2 

and dm5.3 (Wang et al., 2016; Tan et al., 2022).  The narrowed QTL identified by fine mapping 

allowed us to distinguish the boundaries between these three QTL.  With distinct borders between 

the QTL, molecular markers targeting specific QTL can be designed for marker assisted selection, 

and to develop breeding lines that can confer multiple diseases by pyramiding resistance QTL 

from different genetic backgrounds.  

Conclusions 

Resistance to Phytophthora capsici is an agronomically important trait in cucumber but 

currently no resistant commercial varieties are available due to the limited research and intricate 

genetic natures of both the pathogen and the host.  A combination of approaches of QTL-seq and 

associated analyses were used to identify QTL for resistances to Phytophthora fruit rot in 

cucumber.  Multiple QTL were identified for young fruit resistance. The largest effect QTL, 

qPFR5.1, was located on chromosome 5, and narrowed to approximately 1 Mb. A major effect 

QTL for ARR, qPARR3.1, was found at the end of chromosome 3, and a candidate gene identified 

from comparative transcriptomic analyses of cucumber peels.  Additional SNPs associated with 

resistance were discovered from GWAS and XP-GWAS analyses of the USDA cucumber core 

collection.  The close vicinity of the QTL and SNPs identified from multiple analyses 

strengthened the credibility of these findings.  Several of the findings also corresponded with 

previously identified disease resistant hot spots in cucumber.  Collectively, the results of this work 

can provide useful information for future studies to understand mechanisms of resistance to P. 

73 

 
capsici in cucumber and breed for varieties with resistance to Phytophthora fruit rot.  

74 

 
 
 
WORKS CITED 

Alqudah, A. M., Sallam, A., Stephen Baenziger, P., and Börner, A. (2020). GWAS: Fast-

forwarding gene identification and characterization in temperate cereals: lessons from 
barley – A review. J. Adv. Res. 22, 119–135. doi: 10.1016/j.jare.2019.10.013. 

Ando, K., Carr, K. M., Colle, M., Mansfeld, B. N., and Grumet, R. (2015). Exocarp properties 
and transcriptomic analysis of cucumber (Cucumis sativus) fruit expressing age-related 
resistance to Phytophthora capsici. PLOS ONE 10, e0142133. doi: 
10.1371/journal.pone.0142133. 

Ando, K., Carr, K. M., and Grumet, R. (2012). Transcriptome analyses of early cucumber fruit 

growth identifies distinct gene modules associated with phases of development. BMC 
Genomics 13, 518. doi: 10.1186/1471-2164-13-518. 

Ando, K., and Grumet, R. (2006). Evaluation of altered cucumber plant architecture as a means 
to reduce Phytophthora capsici disease incidence on cucumber fruit. J. Amer. Soc. Hort. 
Sci. 131, 491–498. 

Ando, K., Grumet, R., Terpstra, K., and Kelly, J. D. (2007). Manipulation of plant architecture to 

enhance crop disease control. CABI Reviews 2007, 2:026 doi: 
10.1079/PAVSNNR20072026. 

Ando, K., Hammar, S., and Grumet, R. (2009). Age-related resistance of diverse cucurbit fruit to 

infection by Phytophthora capsici. J. Amer. Soc. Hort. Sci. 134, 176–182. doi: 
10.21273/JASHS.134.2.176. 

Barchenger, D. W., Lamour, K. H., and Bosland, P. W. (2018). Challenges and strategies for 
breeding resistance in Capsicum annuum to the multifarious pathogen, Phytophthora 
capsici. Front. Plant Sci. 9:628.  https://doi.org/10.3389/fpls.2018.00628 

Bates, D., Mächler, M., Bolker, B., Walker, S. (2015). Fitting linear mixed-effects models using 

lme4.” J. Stat. Softw. 67, 1–48. doi:10.18637/jss.v067.i01. 

Berg JA, Hermans FWK, Beenders F, Lou L, Vriezen WH, Visser RGF, Bai Y, Schouten HJ. 

2020. Analysis of  QTL DM4.1 for downy mildew resistance in cucumber reveals 
multiple subQTL: a novel RLK as a candidate gene for the most important subQTL. 
Front. Plant Sci. 11: 569876. https://doi.org/10.3389/fpls.2020.569876 

Bolger, A. M., Lohse, M., and Usadel, B. (2014). Trimmomatic: a flexible trimmer for Illumina 
sequence data. Bioinformatics 30, 2114–2120. doi: 10.1093/bioinformatics/btu170. 

Chanock, S. J., Manolio, T., Boehnke, M., Boerwinkle, E., Hunter, D. J., Thomas, G., et al. 

(2007). Replicating genotype–phenotype associations. Nature 447, 655–660. doi: 
10.1038/447655a. 

75 

 
Colle, M., Straley, E. N., Makela, S. B., Hammar, S. A., and Grumet, R. (2014). Screening the 

cucumber plant introduction collection for young fruit resistance to Phytophthora capsici. 
HortScience 49, 244–249. 

Crain, J., Larson, S., Sthapit, S., Jensen, K., Poland, J., Dorn, K., et al. (2023). Genomic insights 
into the NPGS intermediate wheatgrass germplasm collection. Crop Sci. 63, 1381–1396. 
doi: 10.1002/csc2.20944. 

Danecek, P., Bonfield, J. K., Liddle, J., Marshall, J., Ohan, V., Pollard, M. O., et al. (2021). 

Twelve years of SAMtools and BCFtools. Gigascience 10, giab008. doi: 
10.1093/gigascience/giab008. 

DePristo, M. A., Banks, E., Poplin, R., Garimella, K. V., Maguire, J. R., Hartl, C., et al. (2011). 
A framework for variation discovery and genotyping using next-generation DNA 
sequencing data. Nat. Genet. 43, 491–8. doi: 10.1038/ng.806. 

Dunn, A. R., Fry, B. A., Lee, T. Y., Conley, K. D., Balaji, V., Fry, W. E., et al. (2013). 

Transformation of Phytophthora capsici with genes for green and red fluorescent protein 
for use in visualizing plant-pathogen interactions. Aust. Plant Pathol. 42, 583–593. doi: 
10.1007/s13313-013-0222-2. 

Gevens,  A.J., Ando, K., Lamour, K. H., Grumet, R., and Hausbeck, M. K. (2006). A detached 

cucumber fruit method to screen for resistance to Phytophthora capsici and effect of fruit 
age on susceptibility to infection. Plant Dis. 90, 1276–1282. doi: 10.1094/PD-90-1276. 

Granke, L.L., Quesada-Ocampo, L., Lamour, K., and Hausbeck, M.K. (2012). Advances in 
research on Phytophthora capsici on vegetable crops in the United States. Plant Dis. 
95,1588–1600. doi: 10.1094/PDIS-02-12-0211-FE 

Grumet, R., and Colle, M. (2017). Cucumber (Cucumis sativus) breeding line with young fruit 
resistance to infection by Phytophthora capsici. HortScience 52, 922–924. doi: 
10.21273/HORTSCI11423-16. 

Grumet, R., Fei, Z., Levi, A., Mazourek, M., McCreight, J. D., Schultheis, J., et al. (2020). The 
CucCAP project: leveraging applied genomics to improve disease resistance in cucurbit 
crops. Acta Hortic., 91–104. doi: 10.17660/ActaHortic.2020.1294.12. 

Grumet, R., McCreight, J. D., McGregor, C., Weng, Y., Mazourek, M., Reitsma, K., et al. 

(2021). Genetic resources and vulnerabilities of major cucurbit crops. Genes (Basel) 12, 
1222. doi: 10.3390/genes12081222. 

Hausbeck, M. K., and Lamour, K. H. (2004). Phytophthora capsici on vegetable crops: research 

progress and management challenges. Plant Dis. 88, 1292–1303. doi: 
10.1094/PDIS.2004.88.12.1292. 

Huang, S., Li, R., Zhang, Z., Li, L., Gu, X., Fan, W., et al. (2009). The genome of the cucumber, 

76 

 
Cucumis sativus L. Nat. Genet. 41, 1275–1281. doi: 10.1038/ng.475. 

Kiss, E., Oláh, B., Kaló, P., Morales, M., Heckmann, A. B., Borbola, A., et al. (2009). LIN, a 
novel type of U-Box/WD40 Protein, controls early infection by Rhizobia in legumes. 
Plant Physiol. 151, 1239–1249. doi: 10.1104/pp.109.143933. 

Kousik, C.S., Ji, P., Egel, D. S., and Quesada-Ocampo, L. M. (2017). Fungicide rotation 

programs for managing Phytophthora fruit rot of watermelon in Southeastern United 
States. Plant Health Prog. 18, 28–34. doi: 10.1094/PHP-RS-16-0059. 

Kumar, S., Deng, C. H., Molloy, C., Kirk, C., Plunkett, B., Lin-Wang, K., et al. (2022). Extreme-
phenotype GWAS unravels a complex nexus between apple (Malus domestica) red-flesh 
colour and internal flesh browning. Fruit Res. 2, 12. doi: 10.48130/FruRes-2022-0012. 

Latijnhouwers, M., de Wit, P. J. G. M., and Govers, F. (2003). Oomycetes and fungi: similar 

weaponry to attack plants. Trends Microbiol. 11, 462–469. doi: 
10.1016/j.tim.2003.08.002. 

Li, H. (2013). Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. 

http://arxiv.org/abs/1303.3997. 

Li, H., and Durbin, R. (2009). Fast and accurate short read alignment with Burrows-Wheeler 
transform. Bioinformatics 25, 1754–60. doi: 10.1093/bioinformatics/btp324. 

Li, Z., and Xu, Y. (2022). Bulk segregation analysis in the NGS era: a review of its teenage 

years. Plant J. 109, 1355–1374. doi: 10.1111/tpj.15646. 

Li, Z., Zhang, Z., Yan, P., Huang, S., Fei, Z., and Lin, K. (2011). RNA-Seq improves annotation 

of protein-coding genes in the cucumber genome. BMC Genom. 12, 540. doi: 
10.1186/1471-2164-12-540. 

Love, M. I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and 

dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550. doi: 10.1186/s13059-
014-0550-8. 

Lv, J., Qi, J., Shi, Q., Shen, D., Zhang, S., Shao, G., et al. (2012). Genetic diversity and 

population structure of cucumber (Cucumis sativus L.). PLoS ONE 7, e46919. doi: 
10.1371/journal.pone.0046919. 

Mackay, I.J., and Caligari, P.D.S. (2000). Efficiencies of F2 and backcross generations for bulked 

segregant analysis using dominant markers. Crop Sci. 40, 626–630. doi: 
10.2135/cropsci2000.403626x. 

Magwene, P.M., Willis, J.H., Kelly, J.K. (2011) The statistics of bulk segregant analysis using 

next generation sequencing. PLoS Comput. Biol. 7, e1002255. 
https://doi.org/10.1371/journal.pcbi.1002255 

77 

 
Mansfeld, B. N., Colle, M., Kang, Y., Jones, A. D., and Grumet, R. (2017). Transcriptomic and 
metabolomic analyses of cucumber fruit peels reveal a developmental increase in 
terpenoid glycosides associated with age-related resistance to Phytophthora capsici. 
Hortic. Res. 4, 1–11. doi: 10.1038/hortres.2017.22. 

Mansfeld, B. N., Colle, M., Zhang, C., Lin, Y.-C., and Grumet, R. (2020). Developmentally 
regulated activation of defense allows for rapid inhibition of infection in age-related 
resistance to Phytophthora capsici in cucumber fruit. BMC Genomics 21, 628. doi: 
10.1186/s12864-020-07040-9. 

Mansfeld, B. N., and Grumet, R. (2018). QTLseqr: An R Package for Bulk Segregant Analysis 

with Next-Generation Sequencing. Plant Genome 11, 0. doi: 
10.3835/plantgenome2018.01.0006. 

McCreight, J. D., Staub, J. E., Wehner, T. C., and Dhillon, N. P. S. (2013). Gone global: Familiar 

and exotic cucurbits have Asian origins. HortScience 48, 1078–1089. doi: 
10.21273/HORTSCI.48.9.1078. 

McKenna, A., Hanna, M., Banks, E., Sivachenko, A., Cibulskis, K., Kernytsky, A., et al. (2010). 
The Genome Analysis Toolkit: A MapReduce framework for analyzing next-generation 
DNA sequencing data. Genome Res. 20, 1297–1303. doi: 10.1101/gr.107524.110. 

Patro, R., Duggal, G., Love, M.I., Irizarry, R.A., Kingsford, C. (2017). Salmon provides fast and 

bias-aware quantification of transcript expression. Nat. Methods 14, 417–9. 
https://doi.org/10.1038/nmeth.4197. 

Poudel, H. P., Tilhou, N. W., Sanciangco, M. D., Vaillancourt, B., Kaeppler, S. M., Buell, C. R., 
et al. (2021). Genetic loci associated with winter survivorship in diverse lowland 
switchgrass populations. Plant Genome 14, e20159. doi: 10.1002/tpg2.20159. 

Qi, J., Liu, X., Shen, D., Miao, H., Xie, B., Li, X., et al. (2013). A genomic variation map 

provides insights into the genetic basis of cucumber domestication and diversity. Nat. 
Genet. 45, 1510–1515. doi: 10.1038/ng.2801. 

Quesada-Ocampo, L. M., Parada-Rojas, C. H., Hansen, Z., Vogel, G., Smart, C., Hausbeck, M. 

K., et al. (2023). Phytophthora capsici: recent progress on fundamental biology and 
disease management 100 years after its description. Ann. Rev. Phytopath. 61, null. doi: 
10.1146/annurev-phyto-021622-103801. 

Ramos, A., Fu, Y., Michael, V., and Meru, G. (2020). QTL-seq for identification of loci 

associated with resistance to Phytophthora crown rot in squash. Sci. Rep. 10, 5326. doi: 
10.1038/s41598-020-62228-z. 

Rett-Cadman, S., Colle, M., Mansfeld, B., Barry, C. S., Wang, Y., Weng, Y., et al. (2019). QTL 

and transcriptomic analyses implicate cuticle transcription factor SHINE as a source of 
natural variation for epidermal traits in cucumber fruit. Front. Plant Sci. 10. 

78 

 
Rey, T., Chatterjee, A., Buttay, M., Toulotte, J., and Schornack, S. (2015). Medicago truncatula 
symbiosis mutants affected in the interaction with a biotrophic root pathogen. New 
Phytol. 206, 497–500. doi: 10.1111/nph.13233. 

Ries, D., Holtgräwe, D., Viehöver, P., and Weisshaar, B. (2016). Rapid gene identification in 

sugar beet using deep sequencing of DNA from phenotypic pools selected from breeding 
panels. BMC Genomics 17, 236. doi: 10.1186/s12864-016-2566-9. 

Sanogo, S., Lamour, K., Kousik, S., Lozada, D. N., Parada Rojas, C. H., Quesada-Ocampo, L., et 

al. (2022). Phytophthora capsici, 100 years later: Research mile markers from 1922 to 
2022. Phytopathology. doi: 10.1094/PHYTO-08-22-0297-RVW. 

Savory, E. A., Granke, L. L., Quesada-Ocampo, L. M., Varbanova, M., Hausbeck, M. K., and 

Day, B. (2011). The cucurbit downy mildew pathogen Pseudoperonospora cubensis. 
Molec. Plant Pathol. 12, 217–226. doi: 10.1111/j.1364-3703.2010.00670.x. 

Soneson, C., Love, M. I., and Robinson, M. D. (2015). Differential analyses for RNA-seq: 

transcript-level estimates improve gene-level inferences. F1000Res 4, 1521. doi: 
10.12688/f1000research.7563.2. 

Takagi, H., Abe, A., Yoshida, K., Kosugi, S., Natsume, S., Mitsuoka, C., et al. (2013). QTL-seq: 

rapid mapping of quantitative trait loci in rice by whole genome resequencing of DNA 
from two bulked populations. Plant J. 74, 174–183. doi: 10.1111/tpj.12105. 

Tan, J., Wang, Y., Dymerski, R., Wu, Z., and Weng, Y. (2022). Sigma factor binding protein 1 

(CsSIB1) is a putative candidate of the major-effect QTL dm5.3 for downy mildew 
resistance in cucumber (Cucumis sativus). Theor. Appl. Genet. doi: 10.1007/s00122-022-
04212-x. 

Uffelmann, E., Huang, Q. Q., Munung, N. S., de Vries, J., Okada, Y., Martin, A. R., et al. 
(2021). Genome-wide association studies. Nat. Rev. Methods Primers 1, 1–21. doi: 
10.1038/s43586-021-00056-9. 

Van der Auwera, G. A., Carneiro, M. O., Hartl, C., Poplin, R., Del Angel, G., Levy-Moonshine, 
A., et al. (2013). From FastQ data to high confidence variant calls: the Genome Analysis 
Toolkit best practices pipeline. Curr. Protoc. Bioinformatics 43, 11.10.1-33. doi: 
10.1002/0471250953.bi1110s43. 

Vogel, G., Giles, G., Robbins, K. R., Gore, M. A., and Smart, C. D. (2022). Quantitative genetic 

analysis of interactions in the pepper–Phytophthora capsici pathosystem. MPMII 35, 
1018–1033. doi: 10.1094/MPMI-12-21-0307-R. 

Wang, J., and Zhang, Z. (2021). GAPIT Version 3: Boosting power and accuracy for genomic 
association and prediction. GPB 19, 629–640. doi: 10.1016/j.gpb.2021.08.005. 

Wang, X., Bao, K., Reddy, U. K., Bai, Y., Hammar, S. A., Jiao, C., et al. (2018). The USDA 

79 

 
cucumber (Cucumis sativus L.) collection: genetic diversity, population structure, 
genome-wide association studies, and core collection development. Hortic. Res. 5, 1–13. 
doi: 10.1038/s41438-018-0080-8. 

Wang, Y., Bo, K., Gu, X., Pan, J., Li, Y., Chen, J., et al. (2020). Molecularly tagged genes and 

quantitative trait loci in cucumber with recommendations for QTL nomenclature. Hortic. 
Res. 7, 3. doi: 10.1038/s41438-019-0226-3. 

Wang, Y., VandenLangenberg, K., Wehner, T. C., Kraan, P. A. G., Suelmann, J., Zheng, X., et 
al. (2016). QTL mapping for downy mildew resistance in cucumber inbred line WI7120 
(PI 330628). Theor. Appl. Genet. 129, 1493–1505. doi: 10.1007/s00122-016-2719-x. 

Xiao, N., Wu, Y., Pan, C., Yu, L., Chen, Y., Liu, G., et al. (2017). Improving of rice blast 
resistances in Japonica by pyramiding major R genes. Front. Plant Sci. 7.  

Yang, J., Jiang, H., Yeh, C., Yu, J., Jeddeloh, J. A., Nettleton, D., et al. (2015). Extreme-

phenotype genome-wide association study (XP-GWAS): a method for identifying trait‐
associated variants by sequencing pools of individuals selected from a diversity panel. 
Plant J. 84, 587–596. doi: 10.1111/tpj.13029. 

Yin, L., Zhang, H., Tang, Z., Xu, J., Yin, D., Zhang, Z., Yuan, X., Zhu, M., Zhao, S., Li, X., and 
Liu, X. (2021). rMVP: A memory-efficient, visualization-enhanced, and parallel-
accelerated tool for genome-wide association study. GPB 19, 619-628. doi: 
10.1016/j.gpb.2020.10.007. 

Yu, J., Wu, S., Sun, H., Wang, X., Tang, X., Guo, S., et al. (2023). CuGenDBv2: an updated 

database for cucurbit genomics. Nucl. Acids Res. 51, D1457-D1464. doi: 
10.1093/nar/gkac921. 

Zhang, C., Mansfeld, B. N., Lin, Y.-C., and Grumet, R. (2021). Quantitative high-throughput, 

real-time bioassay for plant pathogen growth in vivo. Front.Plant Sci. 12.  

Zou, C., Wang, P., and Xu, Y. (2016). Bulked sample analysis in genetics, genomics and crop 

improvement. Plant Biotech. J. 14, 1941–1955. doi: 10.1111/pbi.12559. 

80 

 
 
 
 
APPENDIX A: Supplementary Materials  

Supplementary Tables (provided in a separate Excel file) 
Sup. Table 2.1. List of cucumber core collection accessions tested for young fruit resistance to 
Phytophthora capsici in GWAS and XP-GWAS analyses.  
Sup. Table 2.2. KASP primer sequences and the location and ref/alt alleles of the targeted SNPs 
for fine mapping qPFR5.1. 
Sup. Table 2.3. Significant SNPs detected in XP-GWAS analysis. Threshold indicates 
significance of the SNPs. 
Sup. Table 2.4. Genes within the qPARR3.1 locus identified as differentially expressed uniquely 
in resistant-aged `Poinsett 76’ fruit. 
Sup. Table 2.5. Polymorphisms detected starting from 2 kb upstream of Csa3G872720 until the 
end of its 3’UTR in four genotypes: ‘Chinese Long 9930’ (ARR-, the reference genome), ‘Gy 
14’ (ARR-), ‘Poinsett 76’ (ARR+) and ‘Vlaspik’ (ARR+). 

Supplementary Figures 
Figure A2.1. SNP-indexes in QTL-seq analyses for young fruit resistance and age-related 
resistance (ARR) in cucumber.  
Figure A2.2. Allelic effect of significant SNPs associated with young fruit resistance identified 
from GWAS analysis. 
Figure A2.3. Disease distribution and individual selection in ARR DH population. 

81 

 
 
 
 
 
A

B

C

Figure A2.1. SNP-indexes in QTL-seq analyses for young fruit resistance and age-related 
resistance (ARR) in cucumber. (A) Young fruit resistance (B) ARR in F2 population, (C) ARR 
in doubled haploid (DH) population. 

82 

 
 
Figure A2.2. Allelic effect of significant SNPs associated with young fruit resistance 
identified from GWAS analysis. 

83 

 
 
 
 
A

B

Figure A2.3. Disease distribution and individual selection in ARR DH population. (A) 
Disease rating distributions within each doubled haploid line. Lines are sorted by mean disease 
rating. Dark lines indicate the median disease rating for each line. Numbers above the median 
indicate the number of fruit tested for that line. In red are the parental lines and the F1. (B) These 
lines were then regrown and phenotyped in a second verification screen in the fall of 2018. The y 
axis represents the mean line disease rating scores from the first screen (Summer) and the x axis 
represents the score in the second screen (Fall). Lines labeled showed consistent phenotypes and 
were selected for bulk segregant analysis. 

84 

 
 
 
CHAPTER 3 

Phenotypic and Genetic Characterization of Morphological Diversity in the Cucumber Core 
Collection 

85 

 
 
Introduction 

Cucumber (Cucumis sativus L.), which is primarily produced for consumption of 

immature fruit, is one of the most highly cultivated vegetable crops worldwide. C. sativus has 

undergone a long history of domestication. Its origin is thought to be in India, China, Burma, 

Thailand, with primary and secondary centers of diversity in India and Southeast Asia, 

respectively (McCreight et al., 2013; Naegele and Wehner, 2016). Following initial 

domestication, cucumber was introduced to China ~2,000 years ago, and was further introduced 

to East Asia (e.g., Japan) around ~100 CE (Lv et al., 2012). It was also dispersed westward to 

Europe and Africa through two routes: Persia into eastern and northern Europe by land in 6th or 7th 

century, and from Persia to western and southern Europe by sea in 10th century (Staub et al., 2008; 

Paris et al., 2012). The plant was later brought to the Americas by Columbus in 1490s.  

Through extensive natural and artificial selection, modern cultivated cucumber shows 

great differences compared to its wild progenitor. For example, cultivated cucumber has fewer 

lateral branches and bears fewer but larger fruits (Harlan, 1992). Many have thicker flesh with 

smaller seed cavity size and no intercentrum hollow, as well as no bitterness (Abbo et al., 2014; 

Weng, 2021). The plants also have larger seeds without dormancy, which can facilitate the 

cultivation process. Despite these general features, cucumber fruit morphology varies 

geographically, reflecting selection based on local environments, consumer preference, and target 

market classes (Weng, 2021; Grumet et al., 2023). The fruits at immature stage are usually crisp 

with underdeveloped seeds for ease of consumption. The major types of cucumber can be 

categorized based on region and usage (fresh or processing). In East Asia, long and slender 

cucumber are preferred and mostly consumed fresh. They can be further classified into North and 

South China types. The North China type, also called “Chinese Long”, typically has long and  

86 

 
tapered fruit covered with dense white spines, while the South China type has cylindrical fruit 

with green and white stripes and fewer black spines (Weng, 2021). North American and Europe 

fresh market cucumbers (slicing cucumbers) have smooth skin and intermediate fruit length (20-

30 cm), while the processing cucumbers (pickling cucumbers) tend to have thicker and warty fruit 

surface, and are generally short in length (5-15 cm). Another major market type is parthenocarpic 

cucumber, typically produced in greenhouses, which includes two classes: long-fruited, about 30-

40 cm long, and Beit Apha (Mediterranean) cucumber, which has shorter fruit (12-15 cm) (Weng, 

2021; Grumet et al., 2023). 

Breeders seeking to develop improved cultivars with crucial traits such as resistances to 

abiotic and biotic stresses to cope with existing and emerging challenges, must also meet market 

standards for fruit quality. These traits include, but are not limited to, fruit size and shape, skin 

and flesh color, spine density, as well as internal traits such as seed cavity size, flesh thickness and 

hollowness. All of these features are crucial when considering the commercial value of a variety. 

Identification of QTL or genes associated these desirable fruit quality traits can facilitate breeding 

efforts to introgress novel resistance or production traits into new cultivars. 

Recent years have seen a rapid increase in genomic tools available for cucumber. The 

USDA-SCRI CucCAP project: Leveraging Applied Genomics to Increase Disease Resistance in 

Cucurbit Crops, has been developing publicly available genetic, genomic , and bioinformatic tools 

for the crops in the Cucurbitaceae family (Grumet et al., 2020). One of the tools developed was a 

cucumber core collection derived from the USDA Agriculture Research Service National Plant 

Germplasm System cucumber collection (NPGS; https://www.ars-grin.gov/Collections#plant-

germplasm) (Wang et al., 2018). The NPGS cucumber collection comprised of 1,314 cultivars, 

landraces, and varieties was sequenced using genotyping-by-sequencing (GBS). The resulting 

87 

 
core collection consists of 388 accessions, which capture > 96% of the allelic diversity in the 

NPGS cucumber collection, and also includes historical cultivars with important agronomic traits 

and disease resistance (Wang et al., 2018). To reduce heterozygosity and heterogeneity, the 

accessions were self-pollinated for 2 or 3 generations. The self-pollinated core collection 

accessions were re-sequenced at 30-40× coverage and used to call SNPs as described by Yu et al., 

(2023). The SNP data are accessible on Cucurbit Genomics Database website (CuGenDB, 

http://cucurbitgenomics.org/v2/).  

In this work, we sought to provide phenotypic and genotypic information about key 

external and internal fruit quality traits present in the cucumber core collection. The resulting 

photographic and quantitative data have been deposited in the publicly accessible CuGenDB, and 

were used to perform Genome-wide association analysis (GWAS) analyses to identify QTL 

associated with the fruit quality traits. With known morphological features and genetic 

characterization of the accessions in the core collection, breeders can select suitable materials that 

fit breeding goals, and further devise efficient and effective strategies to introduce desirable traits 

while maintaining fruit quality standards for the market. 

Materials and Methods 

Plant materials and growth conditions 

The cucumber core collection was selected based on genotyping-by-sequencing (GBS) 

data of United States National Plant Germplasm System (NPGS) collection (Wang et al., 2018). 

To reduce heterozygosity and heterogeneity within the accessions, individuals from each 

accession were self-pollinated for 2 or 3 generations. The self-pollinated core collection lines 

were re-sequenced at 30-40× coverage and used to call SNPs as described by Yu et al. (2023). 

The core collection lines were grown in the field from 2019-2022. The accessions that were tested 

88 

 
are listed in Supplementary Table 3.1. 

The collection was grown at the field of Michigan State University Horticulture Teaching 

and Research Center (HTRC) in 2019-2022 with three plants per accession. Prior to transplanting, 

300 lbs/acre of 19-19-19 fertilizer were applied to the field, and irrigation provided as needed 

throughout the season. The plants were grown on raised black plastic mulch with 0.45-0.6 m 

between plants, depending on the field design each year. The number of accessions planted each 

year varied depending on seed availability. Each accession has 1-4 years of phenotypic data for 

mature fruit and in most cases 1 year of data for young fruit (2021 or 2022). 

Young fruits were harvested at about 5-7 days-post-pollination (dpp), and mature fruits 

were 30-40 dpp. A total of 10-12 fruits were collected from each accession, arranged by size, and 

six representative fruits were selected to photograph. The fruits were rinsed with distilled water to 

remove soil and debris, and then air dried. The fruits were placed a black cloth with a ruler and a 

color correction card for photograph. Data were collected using Fiji software (Fiji Is Just ImageJ, 

Schindelin et al., 2012). Multiple morphological traits were measured from the photographed 

fruits as described below.  

Fruit trait measurements 

Fruit length, diameter, fruit shape index, carpel number, seed cavity, flesh thickness, hollowness, 

curvature, and tapering 

The fruit length was the distance between the blossom and stem ends; when fruits curved, 

length measurement followed the shape of the fruit (Figure 3.1A). Diameter was measured from a 

cross section at the mid-section of the fruit (Figure 3.1B). Fruit shape index was the ratio of fruit 

length/diameter (L/D ratio). Seed cavity size (sc), flesh thickness (ft) and hollowness (h) were 

measured as shown in Figure 3.1B.  

89 

 
Curvature and tapering were measured in young fruits. The curvature measurement was 

adapted from (Clement et al., 2013), which was the ratio of fruit diameter measured at the center 

of the fruit (d) and the distance from the center of the fruit to a line connecting the two ends of the 

fruit (d’) (Figure 3.1C). Higher values indicate greater curvature. Multiple curves were measured 

if a fruit had more than one. Taper was the ratio of the fruit width at the first (q1) and third (q3) 

quarters of a fruit, with higher values indicating more tapering (Figure 3.1D).  

Skin and flesh color, netting, and spine density 

Each photo was color corrected using Chart White Balance plugin in ImageJ 

(https://imagejdocu.list.lu/plugin/color/chart_white_balance/start). RGB color values were 

measured at top, middle and bottom sections for mature fruits and the first and third quarters 

sections for young fruits. The netting was scored for each fruit based on the scale shown in 

Figure 3.1E, where the scale describes the fruit skin from smooth, to lightly and deeply netted 

surface. Spine density was measured on young fruits using ImageJ. The total number of spines on 

each fruit were counted using the cell counter plugin (https://imagej.nih.gov/ij/plugins/cell-

counter.html) and divided by the area of the fruit. The area of the fruit was obtained by 

thresholding color to capture the shape of each fruit.  

Genome-wide association analysis 

The SNP data of the core collection was downloaded from CuGenDBv2 (Yu et al., 

2023).  SNPs were filtered using BCFtools (Danecek et al., 2021) and GATK (Van der Auwera et 

al., 2013) with the following criteria: bi-allelic, GQ scores >20, maximum read depth within one 

standard deviation of the mean read depth, minor allele frequency >0.1, missing rate <20%, 

resulting in 1,179,473 SNPs for association analysis.  Marker-trait association analyses were 

performed using best linear unbiased estimates (BLUEs) calculated from multi-year data. BLUEs 

90 

 
A

l

B

ft

sc

d

h

d

sc

!" =

1
2

(' − )*)

D

q1

q2

& =

'1
'2

! =

#′
#	

d’

d

C

E

Figure 3.1. Illustration of cucumber fruit trait measurement. (A) fruit length (l); (B) internal 
traits: diameter (d, black lines), seed cavity (sc, red lines), hollowness (h, gray line) and flesh 
thickness (ft, blue line); (C) curvature; (D) tapering; (E) fruit netting rating scale. 

were calculated using the R package lme4 (Bates et al., 2015). Association analysis was 

performed using GAPIT 3.0 (Genome Association and Prediction Integrated Tool) with MLM, 

FarmCPU, BLINK, and MLMM models implemented within the software (Wang and Zhang, 

2021). The significance threshold was calculated based on Bonferroni correction (p-value/N, 

where N was the number of SNPs used in the analysis), where the thresholds of adjusted p-

values of 0.05 and 0.01 corresponded to -log10(p) values of 7.37 and 8.06, respectively. 

91 

 
 
 
 
Results 

Phenotypic data in the Cucurbit Genomics Database (CuGenDB) 

Representative mature and young fruits were harvested from 388 accessions during 2019-

2022. The fruits were photographed and phenotyped for a range of fruit quality traits. Examples 

showing diversity for various traits are shown in Figure 3.2. The list of accessions evaluated, 

region of origin and BLUE values for each measured trait are listed in Supplementary Table 3.1. 

The photographic and associated phenotypic data collected in each year have been deposited at 

Cucurbit Genomics Database (CuGenDB, http://cucurbitgenomics.org/v2/), where the 

resequenced genome data of the core collection is also publicly available.  

Phenotypic diversity in the core collection 

Profuse diversity in morphological external and internal traits in mature and young fruits 

was observed in the cucumber core collection (Figure 3.2 and Figure 3.3). Most of the traits were 

normally distributed, such as mature fruit length, diameter, seed cavity size, flesh thickness, and 

tapering, suggesting that there are a wide range of diversity of these traits in the core collection. In 

contrast, several traits showed less variation. Only a few accessions had visibly noticeable 

hollowness at the center of the fruits. Most of the accessions had straight fruit, with low spine 

density and a smooth, non-netted fruit surface (Figure 3.3). Fruit shape related traits and internal 

traits were highly reproducible between years (R = 0.63-0.98) (Supplementary Table 3.2). There 

was a larger range of correlation in color traits among years (R = 0.11-0.89), likely due to 

variation in maturity at the time of harvest. 

Fruit skin color in young fruits primarily varied for degree of greenness, but later 

exhibited a wide range of colors from dark brown, orange, green, to yellow as the fruit matured 

(Figure 3.4A and Figure 3.4B). In order to provide quantitative measurements of color to 

92 

 
facilitate GWAS analysis, RGB values were determined for color traits. When dissecting color 

into RGB descriptive values, low R/G values in young fruits depicted dark green; fruit become 

light green with increasing R/G/B values (Figure 3.4A). In mature fruits, less variation was 

observed in B value compared to R and G values, and low R/G values were associated with darker 

fruit surface. Mature fruit with high B values tended to be orange/yellow while those with low B 

values remained green (Figure 3.4B). Fruit flesh color was predominantly cream and white color, 

with a few showing green or yellow tint (Figure 3.4C). B values in flesh were substantially 

higher than in the peel. 

Correlation between traits 

Correlations among the traits are shown in Figure 3.5. Fruit shape-related traits such as 

young and mature L/D ratios, mature fruit length, and curvature were highly correlated to each 

other (r = 0.83-0.93), and moderately negatively correlated with diameter and seed cavity size (r = 

-0.59-0.72). The longer the fruit, the higher the L/D ratio and greater frequency of curvature. For 

fruit internal traits, diameter was positively correlated with seed cavity size, flesh thickness and 

hollowness (r = 0.56-0.79). Young fruit R and G values were negatively correlated with fruit 

shape traits such as L/D ratios, mature fruit length and curvature (r = 0.42-0.50), and positively 

correlated with seed cavity size and hollowness (r = 0.34-0.44), indicating that fruits with high 

L/D, small seed cavity size and small hollowness tend to have dark green young fruits. Netting is 

a notable trait that largely exists in wild accessions and landraces, which tend to have larger seed 

cavity size and hollowness, and was negatively correlated with fruit shape traits including mature 

fruit length, young and mature L/D ratios and curvature (Figure 3.5).  

Accessions with long fruit shape and high L/D ratios predominantly originated from East 

Asia, where long, thin fruits are preferred (Figure 3.6A). Highly curved fruits were also mostly 

93 

 
found in East Asia accessions. These accessions were more likely to have dark green young fruit 

(lowest R/G values) and turn yellow when mature (high R/G) (Figure 3.6B). While the majority 

of the accessions had fruit with small seed cavity size and smooth skin surface, accessions 

from India and South Asia, which is the center of origin of cucumber, had extensive variation for 

these traits, for example, larger seed cavity size and hollowness, and deeply netted fruit surface. 

These accessions tended to have brown color mature fruits (mid R/G and low B values), and a few 

have flesh color with yellow tone (high R/G and low B values) (Figure 3.6B). Accessions from 

Central/West Asia and Africa also showed moderate variation for netting and hollowness.  

GWAS on morphological external and internal fruit traits 

GWAS was performed on the external and internal traits measured in mature and young 

fruits in the core collection using four models, FarmCPU, BLINK, MLMM, and MLM 

(Supplementary Table 3.3). The Manhattan and QQ plots of FarmCPU model are shown in 

Figure 3.7. The chromosomal locations of significant SNPs identified from the GWAS FarmCPU 

model, and the SNPs that have >10% PVE (phenotypic variance explained) are illustrated in 

Figure 3.8. Though the SNPs identified from the GWAS analysis were widely distributed across 

all chromosomes, QTL for some of the traits were closely clustered (Figure 3.8). For example, on 

chromosome 1 at ~10 Mb, SNPs for several highly correlated fruit size and shape traits, including 

mature fruit length, young fruit L/D, carpel number, and seed cavity size, were closely located. 

Several external fruit traits also mapped to same region on chromosome 1, such as netting, spine 

density, young fruit color R/G values.  

Several QTL identified by GWAS fell within the region of previously identified fruit 

quality QTL and candidate genes (Figure 3.8). The cluster of genes located at ~10 Mb on 

chromosome 1 fell within the consensus QTL for fruit size (FS), and fruit shape index (FSI), 

94 

 
A

B

C

Young Fruit

Skin color
L/D ratio
Tapering
Curvature
Spine density

Mature fruit

External traits
Diameter
Length
L/D ratio
Skin color
Netting

Internal traits
Seed cavity size
Flesh thickness
Hollowness
Flesh color
-

Figure 3.2. Phenotyping fruit quality traits of the cucumber core collection. (A) Examples of 
photographic records of young fruit, mature fruit, and cross sections of the cucumber core 
collection. Graphic and phenotypic data for the full collection can be accessed at Cucurbit 
Genomics Database (CuGenDB) (http://cucurbitgenomics.org/v2/). (B) Geographic distribution of 
accessions in the core collection. (C) Phenotypic traits collected from young and mature fruits in 
the cucumber core collection. 

CsFS1.1 and CsFSI1.1, (Pan et al., 2020). CsFS1.1 was found to contribute to both length and 

diameter growth (Weng et al., 2015). On chromosome 2, significant SNPs for fruit diameter, 

mature fruit L/D, carpel number, and tapering were in the range of CsFS2.1 and CsFSI2.1 at ~10 

Mb. CsFS2.1 was found to contribute to fruit radial growth leading to round fruit shape of 

landrace WI7239 (Pan et al., 2017b). At ~5 Mb on chromosome 3, significant SNPs for flesh  

95 

 
 
 
 
Figure 3.3. Phenotypic distribution of mature and young fruits traits as observed in the 
cucumber core collection. Phenotypes were measured as described in methods. Values of each 
trait were the BLUEs calculated from 1-4 years of data measured for each trait. M – mature fruit; 
Y – young fruit. Photos of each fruit within each plot are to scale.  

thickness and carpel number were located at fth3.1 (Yuan et al., 2008), while at the end of 

chromosome 3, several fruit size and shape related SNPs were also identified in the region of 

CsFS3.2, which was associated with fruit elongation (Weng et al., 2015; Pan et al., 2020). A QTL 

on chromosome 5 identified for spine density overlapped with a previously identified QTL for 

wart density, fws5.1. Though spine density was measured in this study, spine and wart densities 

are closely related traits (Shimomura et al., 2017). At beginning of chromosome 6, significant 

SNPs for seed cavity, mature L/D, carpel number and tapering were colocalized with CsFSI6.1 

(Pan et al., 2020). Consistent with an adjacent pair of fruit shape QTL previously identified at the 

end of chromosome 6 (CsFS6.1 and CsFS6.2) (Pan et al., 2020), two QTL for mature fruit length  

96 

 
 
 
 
were also identified by GWAS in these locations (Supplementary Table 3.3). Lastly, a 

significant SNP for diameter was within the region of CsFS7.2, which was the consensus QTL 

combining multiple fruit shape QTL from Shimomura et al. (2017) (Pan et al., 2020). Although 

not a fruit shape and size trait, significant SNPs for netting were also located in the SNP clusters 

on chromosomes 1, 6, and 7, possibly due to the correlation of netting with other internal traits 

A

Young fruit skin color

B

Mature fruit skin color

C

Mature fruit flesh color

Figure 3.4. Color traits of the cucumber core collection and the distribution of RGB values 
measured for each trait. (A) young fruit skin color; (B) mature fruit skin color; (C) mature fruit 
flesh color.  

97 

 
 
 
and fruit shape traits. Although not detected by FarmCPU, SNPs located in the region of scs1.2, a 

QTL associated with seed cavity size (Yuan et al., 2008), was identified on chromosome 1 in the 

BLINK model (Supplementary Table 3.3). 

Figure 3.5. Correlation (r) matrix of the external and internal traits for young and mature 
fruits in the cucumber core collection. M – mature fruit; Y – young fruit.  

98 

 
 
 
 
 
A

B

Figure 3.6. Distribution of the fruit traits in the core collection based on originated region of 
origin. (A) fruit external and internal traits; (B) color traits. M – mature; Y – young.  

99 

 
 
Figure 3.7. Manhattan and quantile-quantile (QQ) plot of GWAS analysis of the 
morphological traits using FarmCPU model in cucumber core collection. The gray and red 
horizontal lines represent  p-value of 0.05 and 0.01, respectively. 

100 

 
 
x
e
d
n

i

e
p
a
h
s

t
i
u
r
f
e
r
u
t
a
M

x
e
d
n

i

e
p
a
h
s

t
i
u
r
f
g
n
u
o
Y

i

r
e
t
e
m
a
d
t
i
u
r
f
e
r
u
t
a
M

h
t
g
n
e

l

t
i
u
r
f
e
r
u
t
a
M

r
e
b
m
u
n

l

e
p
r
a
C

s
s
e
n
k
c
i
h
t
h
s
e
l
F

e
z
i
s
y
t
i
v
a
c
d
e
e
S

s
w
o

l
l

o
h

l

a
n
r
e
t
n

I

y
t
i
s
n
e
d
e
n
p
S

i

e
r
u
t
a
v
r
u
C

g
n
i
r
e
p
a
T

l

e
u
a
v
R
–
r
o
o
c

l

l

e
u
a
v
G
–
r
o
o
c

l

t
i
u
r
f
g
n
u
o
Y

t
i
u
r
f
g
n
u
o
Y

g
n
i
t
t
e
N

l

e
u
a
v
G
–
r
o
o
c

l

l

e
u
a
v
R
-
r
o
o
c

l

t
i
u
r
f
e
r
u
t
a
M

t
i
u
r
f
e
r
u
t
a
M

l

e
u
a
v
R
–
r
o
o
c
h
s
e
l
F

l

l

e
u
a
v
G
–
r
o
o
c
h
s
e
l
F

l

l

e
u
a
v
B
–
r
o
o
c
h
s
e
l
F

l

L

D

D
L

d

l

C

T

#
C

F

c
S

H

p
S

N

y
R

y
G

m
R

m
G

f
R

f

G

f
B

D
L

C
D

y
R

f
B

c
S
d

l

#
C

C

2
.
3
S
F
s
C

y
R

p
S

f

G

f
B

d

l

D
L

1
.
5
s
w

,
1
.
5
s
w

f

d

l

D
L

#
C

f
B

*

D

L

D
L

d

l

#
C

C

#
C

C
y
R

y
G

f
B

p
S

D
L

f
R

m
G

D

f
B

p
S

1
.
2
I
S
F
s
C

*

,
1
.
2
S
F
s
C

*

C
F

D
L

y
R

y
G

1
.
3
h
t
f

F
H
f

G

f
B

f
R

f
B

f

G

C

T

C

*

p
S

F

C

F

#
C

y
G

m
G
N

c
S

y
R

F
m
G

4
r
h
C

b
M

0
4

0
3

0
2

0
1

f
B

#
C

f
B

m
G

f

G
p
S

y
R

y
G

D

H

p
S

1
.
1
I
S
F
s
C

,
1
.
1
S
F
s
H
C
N*
d

c
S

l

L

d

l

p
S

f

G

C

y
R

1
r
h
C

L

m
G

c
S

f

G

T

T

p
S

T

H

H

#
C

N

d

l

C

L

m
G

2
.
6
S
F
s
C

*
1
.
6
S
F
s
C

f

G

#
C

L

f
B

#
C

2
.
7
S
F
s
C

1
.
6
I
S
F
s
C

H
#
C

f
B

c
S

N

D
L
p
S
F

f
B

c
S

F

F

F

y
G

p
S
N

H

D

N

#
C

C

C

2
r
h
C

3
r
h
C

5
r
h
C

6
r
h
C

7
r
h
C

Figure 3.8. Chromosomal location of the significant SNPs identified in GWAS analysis using 
FarmCPU model for the morphological traits measured in the cucumber core collection. 
SNPs with >10% PVE (phenotypic variance explained) are boxed. The asterisks indicate 
candidate genes identified from prior studies. QTL nomenclature follows Wang et al. (2020b). 

101 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Table 3.1. Previously identified QTL or genes located within 1 Mb of the significant SNPs identified from the GWAS analysis 
using FarmCPU model.  

SNP 

Chr 

Pos (bp) 

P-value 

PVE 
(%) 

QTL/Candidate 
gene 

Gene 

Predicted function 

Refa 

4 

4 

2 
4 

4.60 

Fl4.1 

Fl4.1 

4.39E-11 

3.10 
5.98 

20,280,932 

20,284,337 

7.68E-09 
3.09E-10 

12,445,336 
20,273,647 

CsTRM4/CsTRM5  CsGy2G011350 
CsGy4G016160 

TON1 Recruiting Motif homolog, LONGIFOLIA1 
auxin efflux carrier family protein, PIN-like 6 

Mature L/D 
S2_12445336 
S4_20273647 
Young L/D 
S4_20280932 
Mature length 
S4_20284337 
Flesh thickness 
S2_5438072 
Carpel Number 
S1_10811977 
S1_10811982 
Spine density 
S4_6717620 
Flesh Col - G 
S6_23122636 
a References: 1- Wu et al., 2018, (Xie et al., 2023); 2- (Xing et al., 2023); 3- (Xu et al., 2015); 4- Li et al., 2016; 5- Chen et al., 2016; 6- Kishor et al., 2021 

CsGy4G008590  WD repeat protein (TRANSPARENT TESTA GLABRA1) 

auxin efflux carrier family protein, PIN-like 6 

auxin efflux carrier family protein, PIN-like 6 

SET domain protein-lysine methyltransferase 

CsGy1G014910 
CsGy1G014910 

CLAVATA3 
CLAVATA3 

10,811,977 
10,811,982 

DnaJ Zn finger protein  

1.76E-09 
3.46E-20 

CsCLV3 
CsCLV3 

CsGy4G016160 

CsGy2G006100 

CsGy4G016160 

CsGy6G025020 

14.22 
20.03 

23,122,636 

6,717,620 

5,438,072 

1.29E-15 

1.45E-09 

3.53E-08 

4.34E-09 

CsTTG1 

fth2.1 

10.57 

CsOr 

Fl4.1 

2.69 

1.65 

1.83 

1 
1 

6 

2 

4 

1 
2 

2 

2 

3 

4 

5 

6 

102 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
  
  
  
  
 
  
  
 
Relationship of GWAS identified QTL to prior identified QTL and candidate genes 

Several significant SNPs identified by GWAS were also in close vicinity to prior 

identified candidate genes (Table 3.1 and Figure 3.8). CsCLV3 was located in close proximity to 

two significant SNPs for carpel number at 10.8 Mb on chromosome 1 accounting for 14% and 

20% of the PVE (Supplementary Table 3.3). CsCLV3 is the homolog of Arabidopsis gene 

CLAVATA3 (Li et al., 2016), a regulatory protein involving in controlling cell division and 

differentiation in shoot and floral meristem (Clark et al., 1995; Sablowski, 2011). The expression 

of CsCLV3 was significantly lower in True Lemon, a variety with five carpels, compared to 

WI2757 with three carpels, possibly due to mutation in CsCLV3 in True Lemon (Li et al., 2016). 

A significant SNP for flesh thickness was located near a candidate gene for the flesh thickness 

QTL fth2.1, which encodes a SET domain protein-lysine methyltransferase (PKMT, 

CsGy2G006100). A 4-bp deletion in the promoter region of the PKMT gene resulted in thinner 

flesh (Xu et al., 2015). A candidate gene for fruit shape on chromosome 2, CsTRM4 (also named 

CsTRM5, TONNEAU1 Recruiting Motif5), which is a homolog of tomato SlTRM5, was identified 

for FS2.1 QTL (Wu et al., 2018), and was closely associated with a significant SNP for fruit shape 

(L/D). Knocking out the expression of CsTRM5 (CsaV3_2G013800) repressed cell expansion and 

resulted in shorter fruit length (Xie et al., 2023). For spine density, a significant SNP on 

chromosome 4 is close to a candidate gene CsTTG1 (TRANSPARENT TESTA GLABRA1, 

CsGy4G008590), a TTG1-like homolog in cucumber which encodes a WD repeat protein. 

Mutation of TTG1 in Arabidopsis hampers trichome development (Koornneef, 1981; Walker et 

al., 1999). Upregulating the CsTTG1 gene increased the density of trichomes and spine in 

cucumber fruit (Chen et al., 2016). A tightly clustered set of significant SNPs on chromosome 4 at 

~ 20 Mb was associated with multiple highly correlated fruit shape traits such as fruit length, 

103 

 
diameter, and young and mature L/D ratios, including a SNP for fruit length explaining 11% of 

the variance. An annotated gene (CsGy4G016160) encoding an auxin efflux carrier family protein 

(PIN-LIKES 6 protein) was located within the QTL region and identified as a candidate gene for 

fruit length (Xing et al., 2023). The expression of the gene was negatively correlated with fruit 

length. For flesh color, the CsOr locus (Orange, CsGy6G025020) encoding a DnaJ Zn finger 

protein was close to a significant SNP for flesh color G value on chromosome 6. The CsOr locus 

was identified from CS-B line which had orange flesh due to higher β-carotene content in 

endocarp (Kishor et al., 2021).  

Potential novel genes identified by GWAS analysis 

We also sought to identify potentially novel genes that may influence fruit traits. We 

searched for annotated genes expressed in fruit (according to RNA-seq data accessed in 

CuGenDBv2) that were located within, or within 2 kb upstream, of the GWAS-identified 

significant SNPs . Of the 41 SNPs that fell within or near fruit-expressed genes, 17 were located 

in introns (Supplementary Table 3.4).  Five affected amino acids in exons, but only one caused a 

non-conservative amino acid substitution, glutamine to histidine in an uncharacterized protein.  Of 

the remaining SNPS, two each were within 5’ or 3’ untranslated regions and 15 within 2kb 

upstream of the transcriptional start, potentially affecting gene expression.  

Of particular interest was a set of significant SNPs that were identified for mature fruit 

length and young and mature fruit L/D ratios  located within the coding region of a homolog of 

SCAR (suppressor of cAMP receptor) protein (CsGy4G015630) on chromosome 4 at 20.27 Mb. 

The homolog of SCAR protein in Arabidopsis thaliana is a part of the well-annotated actin 

nucleating promoting complex, WAVE (SCAR/WAVE/WASP [Wiskott–Aldrich syndrome 

protein] family verprolin homologous protein). SCAR/WAVE serves as an activator of the 

104 

 
cytoskeleton-associated actin-related protein complex (ARP2/3) (Deeks and Hussey, 2005; 

Yanagisawa et al., 2013). The role of cytoskeleton involved in cell division and cell morphology 

and organ shape, could potentially influence fruit shape development (Li and Staiger, 2018; 

Livanos and Müller, 2019). However, neither of the identified significant SNPs caused a non-

synonymous amino acid change; one was located in an intron and the second caused a 

conservative amino acid change from phenylalanine to leucine.  

Two additional significant SNPs were located upstream of genes associated with cell 

division factors: one for fruit diameter located 1623 bp upstream of an actin-associated gene on 

chromosome 1, a homolog of Villin-4 (Miears et al., 2018), and one on chromosome 3 for young 

fruit shape located 41bp upstream of a homolog of a gamma-tubulin complex associated protein, 

AUGMIN subunit 1 (Hotta et al., 2012).  Also of potential interest was an additional SNP for 

young fruit shape on chromosome 5 located in the 5’UTR of a homolog of a brassinosteroid-

signaling kinase, BSK-1. Brassinosteroids have been demonstrated to be important for early fruit 

growth (e.g., Fu et al., 2008). 

A second pair of SNPs identified by GWAS for mature fruit color R and G values were 

located close to a gene encoding a DJ-1 homolog D protein (CsGy4G000550) at 0.32 Mb on 

chromosome 4. One SNP was 1.2 kb upstream and the second one was in the 3’ UTR of the gene, 

potentially influencing gene expression. The DJ-1 protein family has been shown to be essential 

in chloroplast development in Arabidopsis thaliana and poplar (Populus trichocarpa) (Lin et al., 

2011; Wang et al., 2023). Disrupting or knockout the expression of DJ-1 protein resulted in albino 

phenotype with malformed chloroplasts (Lin et al., 2011; Wang et al., 2023b).  

Discussion 

Morphological features are crucial when considering breeding for new varieties. New 

105 

 
sources of disease resistance and other valuable traits are often found in landraces or wild 

accessions, many of which have fruit quality traits that are undesirable in commercial markets, for 

example, large seed cavity size and internal hollowness. However, the fruit quality information of 

these accessions is often not available. If an accession possesses a target trait, i.e. disease 

resistance, but distinct fruit quality traits from the market, it might take extra effort and resources 

to introgress the traits and reach acceptable fruit quality. Breeders may prefer to search for 

alternate source materials to pursue breeding goals. Here, the phenotypic data of the core 

collection can provide valuable morphological information for researchers and breeders to 

acquaint them with the plant materials. The 388 accessions in the core collection contain not only 

diverse morphological features, as well as numerous disease resistances and key agronomic traits. 

With available genotypic and phenotypic resources for the core collection, breeders can utilize the 

resources to formulate efficient strategies and develop new varieties.  

Morphological diversity in the core collection 

In this work, morphological external and internal traits were documented and 

characterized (Figure 3.1). Most of the traits exhibit a wide range of diversity in the core 

collection, except for a few that were less variable, including hollowness, carpel number, spine 

density, netting and curvature (Figure 3.2). Though many are landraces and wild cultivars, the 

phenotypes reflect the results of long history of domestication (Wang, 2015; Grumet et al., 2023). 

During the process of domestication (~3,500 years), most cultivated cucumber nowadays is 

phenotypically distinct from the wild ancestor, in addition to fruit traits such as increased flesh 

thickness, smaller seed cavity size, no internal hollowness, and smoother skin surface (Qi et al., 

2013; Weng, 2021). Some distinct phenotypes can also be found in specific regions, showing the 

geographical and cultural selection preference of the crop (Wang, 2015; Grumet et al., 2023). A 

106 

 
noticeable feature was the fruit shape of the accessions originating from East Asia, where these 

accessions predominantly bear long and slender fruit with the average L/D ratios of 8.15 and 6.55 

for young and mature fruit, respectively (Figure 3.6). The tendency of curving fruit was also 

higher in the fruits from this region, probably due to greater capacity for long, slender fruit to 

curve.  

Other evident differences in morphological features were observed in the accessions  from 

India and South Asia. The accessions from these regions tended to have larger seed cavity size 

and hollowness, as well as more heavily netted fruits. These traits are less favorable for 

consumption, and were mostly eliminated during domestication. The fact that India and South 

Asia are the center of origins of cucumber suggests that these accessions preserve the diversity of 

morphology. Additionally, it is known that other than white, yellow and green flesh color, orange 

flesh cucumbers were found in subtropical Xishuangbanna of southwest China and surrounding 

regions (Qi et al., 1983). However, these traits were not observed in the core collection.  

Association among identified fruit trait QTL 

Several SNPs linked to different fruit shape QTL were tightly clustered with each other 

and also located close to, or within, prior QTL or annotated genes (Figure 3.8). A set of 21 

consensus fruit size and shape index QTL identified from multiple populations has been reported 

and summarized (Pan et al., 2020; Wang et al., 2020; Weng, 2021). Among them, some were 

reported to have pleiotropic effects where a single gene or protein influenced multiple traits. For 

example, two fruit shape QTL, CsFS1.4 and CsFS2.3, contribute to both fruit radial growth and 

elongation in the U.S. processing cucumber (Sheng et al., 2020). It is also possible that some traits 

could be polygenic and controlled by multiple closely linked QTL or genes through additive or 

epistatic effects. In addition, several prior identified QTL spanned  relatively large genomic 

107 

 
regions which might include  multiple components that can affect phenotypes. For instance, the 

fruit shape QTL on chromosome 3, CsFS3.2, was associated with fruit elongation but not 

diameter growth (Weng et al., 2015). However, GWAS analysis identified several  significant 

SNP s for mature fruit diameter and young and mature fruit L/D ratios that spanned a 4 Mb 

distance within this QTL region, indicating that the SNPs associated with these traits might be 

associated with distinct genes. Further dissecting the relationship between genomic region and the 

traits could help in pinpointing candidate genes.  

In some cases, different QTL were identified in different stages of fruit development.  For 

instance, SNPs of young fruit L/D were found in the cluster on chromosome 1, but not mature 

fruit L/D, while SNPs for mature fruit L/D were found on chromosomes 2 and 6, but not young 

fruit L/D, suggesting that these QTL may be active in different stages of fruit length and diameter 

growth (Figure 3.8). Detailed analyses have shown that different stages of development are 

associated with peak longitudinal versus lateral expansion and that differences in both cell number 

and cell shape contribute to differences in fruit shape (Colle et al., 2017). Peak L/D ratios 

typically occur early in development and timing varies with genotype (Colle et al. 2017; Pan et 

al., 2020). A prior study of cucumber fruit growth identified several QTL that were active at 

different stages of ovary and fruit development (Weng et al., 2015). Adjacent QTL located near 

the end of chromosome 6, CsFS6.1 and CsFS6.2, both affect length and diameter growth, but 

CsFS6.1 was detected during fruit maturation, while CsFS6.2 was active during development 

from ovary to immature fruit (Weng et al., 2015). Our GWAS analysis also identified two 

adjacent, but distinct, QTL on chromosome 6 corresponding with the locations of CsFS6.1 and 

CsFS6.2. Understanding the dynamic of QTL function in different development stages may also 

be helpful for identifying candidate genes. 

108 

 
From SNPs to candidate genes 

Two SNPs on chromosome 4 (S4_20280932 and S4_20284337) located within the gene 

CsGy4G015630 were identified in multiple fruit shape related traits including mature fruit 

diameter and length, young fruit L/D ratio, and curvature, hinting that the gene associated with 

these SNPs could play a crucial role in fruit development. CsGy4G015630 encodes a homolog of 

an actin-associated SCAR protein that potentially affects cell division and fruit shape. Prior work 

had identified a different candidate gene within the same QTL Fl4.1, CsGy4G016160, encoding a 

homolog of an auxin efflux carrier family protein, PIN-LIKES 6 protein (Xing et al., 2023). 

Auxin has been demonstrated to play an important role in cucumber fruit growth, influencing 

features such as fruit length, neck length, and curvature (e.g., Li et al., 2020; Sharif et al., 2022; 

Wang et al., 2022). While the PIN-LIKES 6 gene was close to the SNPs identified in this study 

(~0.6Mb away), no significant SNPs were detected within the PIN-LIKES 6 gene by GWAS 

analysis by any of the models tested. Examination of the gene expression data provided by Xing 

et al. (2023) for  the 52 genes identified within the interval of the QTL Fl4.1 showed that the 

SCAR gene exhibited a similar expression pattern to the auxin efflux carrier family protein gene. 

Both were significantly downregulated in longer fruit compared to shorter fruit, suggesting that 

both are credible candidate genes.  

Two of the SNPs identified for fruit skin color fell within a gene encoding a homolog of a 

DJ-1 protein, which is involved in chloroplast development. Not surprisingly, numerous prior 

studies have demonstrated  the importance of chloroplasts in fruit skin color, where higher 

chloroplast content led to darker skin color (e.g., Zhou et al., 2015; Liu et al., 2016; Tang et al., 

2018; Wang et al., 2020a). 

While GWAS has been widely used to identify QTL for various traits, there is not a clear 

109 

 
standard for determining causal genes from GWAS analysis and the criteria have been determined 

case by case. Most of the SNPs identified by GWAS are in non-coding and intergenic regions 

(Farh et al., 2015), with difficulty to link to a casual gene. It was reported that the casual genes 

can be located up to 2 Mb away from the SNP, and the closest genes were not necessarily the 

target genes (Brodie et al., 2016). Many factors can affect the precision of GWAS analysis, such 

as accuracy and distribution of phenotypic data, population size and structure, allele frequency of 

traits of interest, and linkage disequilibrium (LD) or haplotype blocks (Alqudah et al., 2020). 

Therefore, many GWAS analyses to date use a combination of criteria, for example, physical 

distance, LD and minor allele frequency, to determine casual variants. In this work, we searched 

for candidate genes by two computational approaches: 1) QTL and genes that were previously 

identified from literature and located within 1 Mb from the SNPs; and 2) inspected the genes that 

were located within 2 kb upstream of the significant SNPs. We also took advantage of extensively 

curated RNA-seq data available in CuGenDB (http://cucurbitgenomics.org/v2/) to identify those 

that are expressed in cucumber fruits (Supplementary Table 3.4). While these approaches 

resulted in a more condensed list of candidate genes, ultimately, it will be necessary to directly 

verify the function and expression of the candidate genes. 

Using RGB values in phenotyping color-related traits  

Fruit skin and flesh color in this study were measured using calibrated RGB values 

(Figure 3.4). The use of RGB values can provide advantages relative to categorical color rating 

scales or analysis of pigment content, for example, providing  digitally accessible image data, 

quantitative measurement compared to visual rating, and reduced cost relative to biochemical 

analyses of chlorophyll or carotenoid content (Jiao et al., 2017; Hao et al., 2018; Kishor et al., 

2021; Sun et al., 2023). RGB values have been widely used in various high-throughput 

110 

 
phenotyping research such as surveying vegetation or monitoring plant growth (Chen et al., 2020; 

Sánchez-Sastre et al., 2020; Han et al., 2021). 

More intense color both in young and mature fruits was associated with low R and G 

values, , likely representing increased chloroplast (and subsequent chromoplast) content. 

Similarly, flesh tissue ,which was predominantly white and lacking chloroplasts, had higher R and 

G values. Mature fruits with lower G values tended to be brown or orange, instead of green; while 

in flesh color, lower G values corresponded to increased yellow color, suggesting those might 

have higher carotenoid content. Multiple models have been proposed to scrutinize the relationship 

between RGB values and traits of interest (Zakaluk and Sri Ranjan, 2006; Beamish et al., 2018; 

Alves et al., 2022; Olivoto, 2022). Consistent with our observations, research tracking seasonal 

vegetation changes in low-Arctic tundra found that G value highly reflects chlorophyll content, 

while R-based index reflected chlorophyll to carotenoid ratio moderately (Beamish et al., 2018).  

Conclusion 

This work provided morphological characterization of fruit quality traits in the cucumber 

core collection. The multi-year phenotypic and photographic data have been deposited the in 

CuGenDB website. In addition, GWAS analysis was performed for the  morphological traits 

measured and provided a refined list of candidate genes that could be useful for future research. In 

many cases, QTL identified from this study overlapped with prior QTL mapping studies, 

providing confidence in the analyses and further support for  QTL and candidate genes in the 

literature. Notably, most of the morphological QTL identified in cucumber to date, are from QTL 

mapping using bi-parental populations. Only a few have been performed using diversity panels 

for a limited number of traits: green fruit flesh (Bo et al., 2019b), spine density (Bo et al., 2019a), 

pericarp color (Huang et al., 2022b), and fruit skin waxiness (Huang et al., 2022a). Bi-parental 

111 

 
populations derived from two parents with extreme phenotypes provide a quick and specific 

method to explore QTL underlying a trait. However, the genetic variation is limited to the two 

parents and the size of QTL is affected by population structure and size (Mackay and Caligari, 

2000; Li and Xu, 2022). A diversity panel composed of a set unrelated individuals can provide a 

greater range of diversity, but requires higher expense of genotyping and phenotyping, as well as 

lower ability to detect rare alleles (Alqudah et al., 2020; Uffelmann et al., 2021). In this work we 

were able to take advantage of the established cucumber core collection and the associated 

resequencing data providing the necessary genetic data needed for genomic analysis. The 

collected fruit quality phenotypic documentation and photographic data will provide an additional 

valuable resources for breeding research. 

112 

 
 
 
WORKS CITED 

Abbo, S., Pinhasi van-Oss, R., Gopher, A., Saranga, Y., Ofner, I., and Peleg, Z. (2014). Plant 

domestication versus crop evolution: a conceptual framework for cereals and grain 
legumes. Trends in Plant Science 19, 351–360. doi: 10.1016/j.tplants.2013.12.002. 

Alqudah, A. M., Sallam, A., Stephen Baenziger, P., and Börner, A. (2020). GWAS: Fast-

forwarding gene identification and characterization in temperate Cereals: lessons from 
Barley – A review. Journal of Advanced Research 22, 119–135. doi: 
10.1016/j.jare.2019.10.013. 

Alves, K. S., Guimarães, M., Ascari, J. P., Queiroz, M. F., Alfenas, R. F., Mizubuti, E. S. G., et 

al. (2022). RGB-based phenotyping of foliar disease severity under controlled conditions. 
Trop. plant pathol. 47, 105–117. doi: 10.1007/s40858-021-00448-y. 

Bates, D., Mächler, M., Bolker, B., and Walker, S. (2015). Fitting Linear Mixed-Effects Models 
Using lme4. Journal of Statistical Software 67, 1–48. doi: 10.18637/jss.v067.i01. 

Beamish, A. L., Coops, N. C., Hermosilla, T., Chabrillat, S., and Heim, B. (2018). Monitoring 

pigment-driven vegetation changes in a low-Arctic tundra ecosystem using digital 
cameras. Ecosphere 9, e02123. doi: 10.1002/ecs2.2123. 

Bo, K., Miao, H., Wang, M., Xie, X., Song, Z., Xie, Q., et al. (2019a). Novel loci fsd6.1 and 

Csgl3 regulate ultra-high fruit spine density in cucumber. Theor Appl Genet 132, 27–40. 
doi: 10.1007/s00122-018-3191-6. 

Bo, K., Wei, S., Wang, W., Miao, H., Dong, S., Zhang, S., et al. (2019b). QTL mapping and 

genome-wide association study reveal two novel loci associated with green flesh color in 
cucumber. BMC Plant Biology 19, 243. doi: 10.1186/s12870-019-1835-6. 

Brodie, A., Azaria, J. R., and Ofran, Y. (2016). How far from the SNP may the causative genes 

be? Nucleic Acids Res 44, 6046–6054. doi: 10.1093/nar/gkw500. 

Chen, C., Yin, S., Liu, X., Liu, B., Yang, S., Xue, S., et al. (2016). The WD-Repeat Protein 

CsTTG1 Regulates Fruit Wart Formation through Interaction with the Homeodomain-
Leucine Zipper I Protein Mict1. Plant Physiol 171, 1156–1168. doi: 
10.1104/pp.16.00112. 

Chen, Z., Wang, F., Zhang, P., Ke, C., Zhu, Y., Cao, W., et al. (2020). Skewed distribution of 
leaf color RGB model and application of skewed parameters in leaf color description 
model. Plant Methods 16, 23. doi: 10.1186/s13007-020-0561-2. 

Clark, S. E., Running, M. P., and Meyerowitz, E. M. (1995). CLAVATA3 is a specific regulator 
of shoot and floral meristem development affecting the same processes as CLAVATA1. 
Development 121, 2057–2067. doi: 10.1242/dev.121.7.2057. 

113 

 
Clement, J., Novas, N., Gazquez, J.-A., and Manzano-Agugliaro, F. (2013). An active contour 
computer algorithm for the classification of cucumbers. Computers and Electronics in 
Agriculture 92, 75–81. doi: 10.1016/j.compag.2013.01.006. 

Colle, M., Weng, Y., Kang, Y., Ophir, R., Sherman, A., and Grumet, R. (2017). Variation in 
cucumber (Cucumis sativus L.) fruit size and shape results from multiple components 
acting pre-anthesis and post-pollination. Planta. doi: 10.1007/s00425-017-2721-9. 

Danecek, P., Bonfield, J. K., Liddle, J., Marshall, J., Ohan, V., Pollard, M. O., et al. (2021). 

Twelve years of SAMtools and BCFtools. Gigascience 10, giab008. doi: 
10.1093/gigascience/giab008. 

Deeks, M. J., and Hussey, P. J. (2005). Arp2/3 and SCAR: plants move to the fore. Nat Rev Mol 

Cell Biol 6, 954–964. doi: 10.1038/nrm1765. 

Farh, K. K.-H., Marson, A., Zhu, J., Kleinewietfeld, M., Housley, W. J., Beik, S., et al. (2015). 

Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature 518, 
337–343. doi: 10.1038/nature13835. 

Fu, F. Q., Mao, W. H., Shi, K., Zhou, Y. H., Asami, T., and Yu, J. Q. (2008). A role of 

brassinosteroids in early fruit development in cucumber. J Exp Bot 59, 2299–2308. doi: 
10.1093/jxb/ern093. 

Grumet, R., Fei, Z., Levi, A., Mazourek, M., McCreight, J. D., Schultheis, J., et al. (2020). The 
CucCAP project: leveraging applied genomics to improve disease resistance in cucurbit 
crops. Acta Hortic., 91–104. doi: 10.17660/ActaHortic.2020.1294.12. 

Grumet, R., Lin, Y.-C., Rett-Cadman, S., and Malik, A. (2023). Morphological and Genetic 
Diversity of Cucumber (Cucumis sativus L.) Fruit Development. Plants 12, 23. doi: 
10.3390/plants12010023. 

Han, G. D., Jang, G., Kim, J., Kim, D.-W., Rodrogues, R., Kim, S.-H., et al. (2021). RGB 

images-based vegetative index for phenotyping kenaf (Hibiscus cannabinus L.). PLOS 
ONE 16, e0256978. doi: 10.1371/journal.pone.0256978. 

Hao, N., Du, Y., Li, H., Wang, C., Wang, C., Gong, S., et al. (2018). CsMYB36 is involved in the 
formation of yellow green peel in cucumber (Cucumis sativus L.). Theor Appl Genet 131, 
1659–1669. doi: 10.1007/s00122-018-3105-7. 

Harlan, J. R. ed. (1992). Crops & Man. Madison, WI, USA: American Society of Agronomy, 

Crop Science Society of America doi: 10.2135/1992.cropsandman. 

Hotta, T., Kong, Z., Ho, C.-M. K., Zeng, C. J. T., Horio, T., Fong, S., et al. (2012). 

Characterization of the Arabidopsis Augmin Complex Uncovers Its Critical Function in 
the Assembly of the Acentrosomal Spindle and Phragmoplast Microtubule Arrays. Plant 
Cell 24, 1494–1509. doi: 10.1105/tpc.112.096610. 

114 

 
Huang, H., Du, Y., Long, Z., Li, Y., Kong, W., Wang, H., et al. (2022a). Fine mapping of a 

novel QTL CsFSG1 for fruit skin gloss in cucumber (Cucumis sativus L.). Mol Breed 42, 
25. doi: 10.1007/s11032-022-01291-y. 

Huang, H., Yang, Q., Zhang, L., Kong, W., Wang, H., Wei, A., et al. (2022b). Genome-wide 
association analysis reveals a novel QTL CsPC1 for pericarp color in cucumber. BMC 
Genomics 23, 383. doi: 10.1186/s12864-022-08606-5. 

Jiao, J., Liu, H., Liu, J., Cui, M., Xu, J., Meng, H., et al. (2017). Identification and functional 

characterization of APRR2 controlling green immature fruit color in cucumber (Cucumis 
sativus L.). Plant Growth Regul 83, 233–243. doi: 10.1007/s10725-017-0304-1. 

Kishor, D. S., Lee, H.-Y., Alavilli, H., You, C.-R., Kim, J.-G., Lee, S.-Y., et al. (2021). 

Identification of an Allelic Variant of the CsOr Gene Controlling Fruit Endocarp Color in 
Cucumber (Cucumis sativus L.) Using Genotyping-By-Sequencing (GBS) and Whole-
Genome Sequencing. Frontiers in Plant Science 12. doi: 10.3389/fpls.2021.802864. 

Koornneef, M. (1981). The complex syndrome of ttg mutants. Arbid. Inf. Serv. 18, 45–51. 

Li, J., and Staiger, C. J. (2018). Understanding Cytoskeletal Dynamics During the Plant Immune 

Response. Annu Rev Phytopathol 56, 513–533. doi: 10.1146/annurev-phyto-080516-
035632. 

Li, S., Pan, Y., Wen, C., Li, Y., Liu, X., Zhang, X., et al. (2016). Integrated analysis in bi-

parental and natural populations reveals CsCLAVATA3 (CsCLV3) underlying carpel 
number variations in cucumber. Theor Appl Genet 129, 1007–1022. doi: 10.1007/s00122-
016-2679-1. 

Li, S., Wang, C., Zhou, X., Liu, D., Liu, C., Luan, J., et al. (2020). The curvature of cucumber 

fruits is associated with spatial variation in auxin accumulation and expression of a 
YUCCA biosynthesis gene. Hortic Res 7, 1–12. doi: 10.1038/s41438-020-00354-5. 

Li, Z., and Xu, Y. (2022). Bulk segregation analysis in the NGS era: a review of its teenage 

years. The Plant Journal 109, 1355–1374. doi: 10.1111/tpj.15646. 

Lin, J., Nazarenus, T. J., Frey, J. L., Liang, X., Wilson, M. A., and Stone, J. M. (2011). A Plant 

DJ-1 Homolog Is Essential for Arabidopsis thaliana Chloroplast Development. PLOS 
ONE 6, e23731. doi: 10.1371/journal.pone.0023731. 

Liu, H., Jiao, J., Liang, X., Liu, J., Meng, H., Chen, S., et al. (2016). Map-based cloning, 

identification and characterization of the w gene controlling white immature fruit color in 
cucumber (Cucumis sativus L.). Theor Appl Genet 129, 1247–1256. doi: 10.1007/s00122-
016-2700-8. 

Livanos, P., and Müller, S. (2019). Division Plane Establishment and Cytokinesis. Annu Rev 

Plant Biol 70, 239–267. doi: 10.1146/annurev-arplant-050718-100444. 

115 

 
Lv, J., Qi, J., Shi, Q., Shen, D., Zhang, S., Shao, G., et al. (2012). Genetic Diversity and 

Population Structure of Cucumber (Cucumis sativus L.). PLoS ONE 7, e46919. doi: 
10.1371/journal.pone.0046919. 

Mackay, I. j., and Caligari, P. d. s. (2000). Efficiencies of F2 and Backcross Generations for 

Bulked Segregant Analysis Using Dominant Markers. Crop Science 40, 626–630. doi: 
10.2135/cropsci2000.403626x. 

McCreight, J. D., Staub, J. E., Wehner, T. C., and Dhillon, N. P. S. (2013). Gone Global: 

Familiar and Exotic Cucurbits Have Asian Origins. HortScience 48, 1078–1089. doi: 
10.21273/HORTSCI.48.9.1078. 

Miears, H. L., Gruber, D. R., Horvath, N. M., Antos, J. M., Young, J., Sigurjonsson, J. P., et al. 

(2018). Plant Villin Headpiece Domain Demonstrates a Novel Surface Charge Pattern 
and High Affinity for F-Actin. Biochemistry 57, 1690–1701. doi: 
10.1021/acs.biochem.7b00856. 

Naegele, R. P., and Wehner, T. C. (2016). “Genetic Resources of Cucumber,” in Plant Genetics 

and Genomics: Crops and Models, 1–26. doi: 10.1007/7397_2016_15. 

Olivoto, T. (2022). Lights, camera, pliman! An R package for plant image analysis. Methods in 

Ecology and Evolution 13, 789–798. doi: 10.1111/2041-210X.13803. 

Pan, Y., Qu, S., Bo, K., Gao, M., Haider, K. R., and Weng, Y. (2017). QTL mapping of 
domestication and diversifying selection related traits in round-fruited semi-wild 
Xishuangbanna cucumber (Cucumis sativus L. var. xishuangbannanesis). Theoretical and 
Applied Genetics. doi: 10.1007/s00122-017-2908-2. 

Pan, Y., Wang, Y., McGregor, C., Liu, S., Luan, F., Gao, M., et al. (2020). Genetic architecture 

of fruit size and shape variation in cucurbits: a comparative perspective. Theor Appl 
Genet 133, 1–21. doi: 10.1007/s00122-019-03481-3. 

Paris, H. S., Daunay, M.-C., and Janick, J. (2012). Occidental diffusion of cucumber (Cucumis 
sativus) 500-1300 CE: two routes to Europe. Annals of Botany 109, 117–126. doi: 
10.1093/aob/mcr281. 

Qi, C. Z., Yuan, Z. Z., and Li, Y. X. (1983). A new type of cucumber. Cucumis sativus L. var. 

xishuang-bannanesis. 

Qi, J., Liu, X., Shen, D., Miao, H., Xie, B., Li, X., et al. (2013). A genomic variation map 

provides insights into the genetic basis of cucumber domestication and diversity. Nature 
Genetics 45, 1510–1515. doi: 10.1038/ng.2801. 

Sablowski, R. (2011). Plant stem cell niches: from signalling to execution. Current Opinion in 

Plant Biology 14, 4–9. doi: 10.1016/j.pbi.2010.08.001. 

Sánchez-Sastre, L. F., Alte da Veiga, N. M. S., Ruiz-Potosme, N. M., Carrión-Prieto, P., Marcos-
Robles, J. L., Navas-Gracia, L. M., et al. (2020). Assessment of RGB Vegetation Indices 

116 

 
to Estimate Chlorophyll Content in Sugar Beet Leaves in the Final Cultivation Stage. 
AgriEngineering 2, 128–149. doi: 10.3390/agriengineering2010009. 

Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, T., et al. 

(2012). Fiji: an open-source platform for biological-image analysis. Nat Methods 9, 676–
682. doi: 10.1038/nmeth.2019. 

Sharif, R., Su, L., Chen, X., and Qi, X. (2022). Involvement of auxin in growth and stress 

response of cucumber. V 2, 1–9. doi: 10.48130/VR-2022-0013. 

Sheng, Y., Pan, Y., Li, Y., Yang, L., and Weng, Y. (2020). Quantitative trait loci for fruit size 

and flowering time-related traits under domestication and diversifying selection in 
cucumber (Cucumis sativus). Plant Breeding 139, 176–191. doi: 10.1111/pbr.12754. 

Shimomura, K., Fukino, N., Sugiyama, M., Kawazu, Y., Sakata, Y., and Yoshioka, Y. (2017). 
Quantitative trait locus analysis of cucumber fruit morphological traits based on image 
analysis. Euphytica 213, 138. doi: 10.1007/s10681-017-1926-0. 

Staub, J. E., Robbins, M. D., and Wehner, T. C. (2008). “Cucumber,” in Vegetables I: 

Asteraceae, Brassicaceae, Chenopodicaceae, and Cucurbitaceae (New York, NY: 
Springer New York), 241–282. doi: 10.1007/978-0-387-30443-4_8. 

Sun, W., Li, X., Huang, H., Wei, J., Zeng, F., Huang, Y., et al. (2023). Mutation of CsARC6 

affects fruit color and increases fruit nutrition in cucumber. Theor Appl Genet 136, 111. 
doi: 10.1007/s00122-023-04337-7. 

Tang, H.-Y., Dong, X., Wang, J.-K., Xia, J.-H., Xie, F., Zhang, Y., et al. (2018). Fine Mapping 

and Candidate Gene Prediction for White Immature Fruit Skin in Cucumber (Cucumis 
sativus L.). International Journal of Molecular Sciences 19, 1493. doi: 
10.3390/ijms19051493. 

Uffelmann, E., Huang, Q. Q., Munung, N. S., de Vries, J., Okada, Y., Martin, A. R., et al. 
(2021). Genome-wide association studies. Nat Rev Methods Primers 1, 1–21. doi: 
10.1038/s43586-021-00056-9. 

Van der Auwera, G. A., Carneiro, M. O., Hartl, C., Poplin, R., Del Angel, G., Levy-Moonshine, 
A., et al. (2013). From FastQ data to high confidence variant calls: the Genome Analysis 
Toolkit best practices pipeline. Current protocols in bioinformatics 43, 11.10.1-33. doi: 
10.1002/0471250953.bi1110s43. 

Walker, A. R., Davison, P. A., Bolognesi-Winfield, A. C., James, C. M., Srinivasan, N., 

Blundell, T. L., et al. (1999). The TRANSPARENT TESTA GLABRA1 locus, which 
regulates trichome differentiation and anthocyanin biosynthesis in Arabidopsis, encodes a 
WD40 repeat protein. Plant Cell 11, 1337–1350. 

Wang, J., and Zhang, Z. (2021). GAPIT Version 3: Boosting Power and Accuracy for Genomic 
Association and Prediction. Genomics, Proteomics & Bioinformatics 19, 629–640. doi: 
10.1016/j.gpb.2021.08.005. 

117 

 
Wang, M., Chen, L., Liang, Z., He, X., Liu, W., Jiang, B., et al. (2020a). Metabolome and 

transcriptome analyses reveal chlorophyll and anthocyanin metabolism pathway 
associated with cucumber fruit skin color. BMC Plant Biology 20, 386. doi: 
10.1186/s12870-020-02597-9. 

Wang, X., Bao, K., Reddy, U. K., Bai, Y., Hammar, S. A., Jiao, C., et al. (2018). The USDA 

cucumber (Cucumis sativus L.) collection: genetic diversity, population structure, 
genome-wide association studies, and core collection development. Hortic Res 5, 1–13. 
doi: 10.1038/s41438-018-0080-8. 

Wang, X., Shao, C., Liu, L., Wang, Y., An, Y., Li, H., et al. (2023). PtrDJ1C, an atypical 

member of the DJ-1 superfamily, is essential for early chloroplast development and lignin 
deposition in poplar. Horticultural Plant Journal. doi: 10.1016/j.hpj.2023.05.011. 

Wang, Y., Bo, K., Gu, X., Pan, J., Li, Y., Chen, J., et al. (2020b). Molecularly tagged genes and 
quantitative trait loci in cucumber with recommendations for QTL nomenclature. 
Horticulture Research 7, 3. doi: 10.1038/s41438-019-0226-3. 

Wang, Z., Zhou, Z., Wang, L., Yan, S., Cheng, Z., Liu, X., et al. (2022). The CsHiiEC1-
CsOVATE module contributes to fruit neck length variation via modulating auxin 
biosynthesis in cucumber. Proceedings of the National Academy of Sciences 119, 
e2209717119. doi: 10.1073/pnas.2209717119. 

Weng, Y. (2021). “Cucumis sativus Chromosome Evolution, Domestication, and Genetic 
Diversity: Implications for Cucumber Breeding,” in Plant Breeding Reviews, ed. I. 
Goldman (Wiley), 79–111. doi: 10.1002/9781119717003.ch4. 

Weng, Y., Colle, M., Wang, Y., Yang, L., Rubinstein, M., Sherman, A., et al. (2015). QTL 

mapping in multiple populations and development stages reveals dynamic quantitative 
trait loci for fruit size in cucumbers of different market classes. Theoretical and Applied 
Genetics 128, 1747–1763. doi: 10.1007/s00122-015-2544-7. 

Wu, S., Zhang, B., Keyhaninejad, N., Rodríguez, G. R., Kim, H. J., Chakrabarti, M., et al. 

(2018). A common genetic mechanism underlies morphological diversity in fruits and 
other plant organs. Nat Commun 9, 4734. doi: 10.1038/s41467-018-07216-8. 

Xie, Y., Liu, X., Sun, C., Song, X., Li, X., Cui, H., et al. (2023). CiTRM5 regulates fruit shape 

via mediating cell division direction and cell expansion in cucumber. Horticulture 
Research 10, uhad007. doi: 10.1093/hr/uhad007. 

Xing, Y., Cao, Y., Ma, Y., Wang, F., Xin, S., and Zhu, W. (2023). QTL mapping and 

transcriptomic analysis of fruit length in cucumber. Frontiers in Plant Science 14.  

Xu, X., Lu, L., Zhu, B., Xu, Q., Qi, X., and Chen, X. (2015). QTL mapping of cucumber fruit 

flesh thickness by SLAF-seq. Sci Rep 5, 15829. doi: 10.1038/srep15829. 

Yanagisawa, M., Zhang, C., and Szymanski, D. B. (2013). ARP2/3-dependent growth in the 
plant kingdom: SCARs for life. Front Plant Sci 4, 166. doi: 10.3389/fpls.2013.00166. 

118 

 
Yu, J., Wu, S., Sun, H., Wang, X., Tang, X., Guo, S., et al. (2023). CuGenDBv2: an updated 

database for cucurbit genomics. Nucleic Acids Research 51, D1457–D1464. doi: 
10.1093/nar/gkac921. 

Yuan, X. J., Pan, J. S., Cai, R., Guan, Y., Liu, L. Z., Zhang, W. W., et al. (2008). Genetic 
mapping and QTL analysis of fruit and flower related traits in cucumber (Cucumis 
sativus L.) using recombinant inbred lines. Euphytica 164, 473–491. doi: 
10.1007/s10681-008-9722-5. 

Zakaluk, R., and Sri Ranjan, R. (2006). Predicting leaf water potential of potato using spectral 

reflectance indices. doi: 10.13031/2013.22139. 

Zhou, Q., Wang, S., Hu, B., Chen, H., Zhang, Z., and Huang, S. (2015). An ACCUMULATION 
AND REPLICATION OF CHLOROPLASTS 5 gene mutation confers light green peel in 
cucumber. Journal of Integrative Plant Biology 57, 936–942. doi: 10.1111/jipb.12355. 

119 

 
 
 
 
APPENDIX B: Supplementary Materials  

Supplementary Tables (provided in a separate Excel file) 
Sup. Table 3.1. The list of accessions n the core collection and the morphological phenotypic 
data measured in 2019-2022.  
Sup. Table 2.2. Correlation between 2019-2022 of mature fruit external and internal traits. 
Sup. Table 2.3. Previously identified QTL or candidate genes located within 1 Mb of the 
significant SNPs identified from the GWAS analysis in all models (FarmCPU, BLINK, MLMM, 
MLM) for morphological traits measured. 
Sup. Table 2.4. Cucumber fruit trait SNPs identified by GWAS (FarmCPU) within or near 
(<2kb upstream) fruit-expressed annotated genes (Gy v.2.1; CuGenDB v.2). 

120 

 
 
 
CHAPTER 4 

Conclusion and Future Directions 

121 

 
 
Phytophthora fruit rot caused by the oomycete pathogen, Phytophthora capsici, is a severe 

disease that causes loss for pickling cucumber growers. The intricate genetic natures of both the 

pathogen and the host hampers the development of resistant cucumber varieties. With the recent 

development of advanced genetic and genomics tools, we were able to use multiple genomic 

approaches and different types of populations to understand the genetics behind the resistance in 

cucumber and to develop molecular markers and genomic information for future breeding efforts 

for resistance to P. capsici. 

We studied young fruit resistance, which is observed during rapid fruit growth prior to 

harvest. First, we utilized a doubled haploid resistant parent derived from the resistant breeding 

line (MSU 109483-53) and developed bi-parental populations for QTL-seq analysis. The results 

confirmed that the young fruit resistance is quantitative trait, with the largest effect QTL, 

qPFR5.1, located on chromosome 5. The effect of qPFR5.1 was additionally verified in a fresh 

market cucumber variety. Then we fine mapped the region to ~1 Mb with KASP markers from 

two F3 populations in two seasons and environments. The QTL position and the sequence of 

flanking makers are available for future breeding work to introgress the QTL into existing 

breeding materials or varieties. It is also important to pay heed to the fact that this is a quantitative 

trait. Though qPFR5.1 had the largest effect, additive or epistatic interactions with other QTL 

should be considered. Two other QTL on chromosomes 1 and 6 were also identified from QTL-

seq analysis, however, were not pursued due to little phenotypic effect or failure to develop 

markers around the QTL region. Understanding the genetic mechanisms of these QTL may shed 

light on a more comprehensive view of young fruit resistance.  

The cucumber core collection provides a wide range of diversity that is useful to mine for 

valuable traits existing in natural populations. With multiple years of data collected, we performed 

122 

 
two types of association analysis: GWAS and XP-GWAS, to search of SNPs that are linked to 

young fruit resistance. Several SNPs were overlapped or located close to the QTL identified from 

the QTL-seq study, as well as QTL that were previously identified for resistances to other 

diseases, such as downy mildew, powdery mildew, and gummy stem blight. These findings were 

consistent with previously observed hot spots for disease resistance on multiple chromosomes. 

Additionally, several accessions consistently had lower disease ratings with minor symptom 

development and can be used as additional sources of P. capsici resistance. Interestingly, the three 

most resistant accessions originate from different regions (Bhutan, China, and the United States) 

and have distinct fruit morphology, suggesting that the resistance to P. capsici may be preserved 

selectively in some accessions around the world but are yet to be discovered. The SNPs reported 

in this work can be informative when choosing materials from the core collection as materials to 

breed for resistances. Collectively, the results of this work can provide useful information for 

future studies to understand mechanisms of resistance to P. capsici in cucumber and breed for 

varieties with resistance to Phytophthora fruit rot. 

In addition to resistance to P. capsici, we also characterized morphological diversity in the 

cucumber core collection. The importance of cucumber fruit quality traits is greatly reflected in its 

commercial value. As numerous abiotic and biotic challenges arise, breeders and researchers often 

seek novel sources of valuable traits in diverse germplasm. However, many sources of traits of 

interest reside in materials with feral phenotypes which are undesirable in today’s market. 

Therefore, we assessed multiple crucial fruit quality traits in the cucumber core collection. The 

photographic documentation and quantitative data for these traits are available on the CuGenDB 

website. This publicly accessible information provides a quick glimpse of the diverse phenotypes 

in the core collection, and breeders can scheme breeding strategies accordingly. Selecting 

123 

 
appropriate breeding materials can possibly achieve breeding goals with shorter breeding 

generations.  

We also performed GWAS analyses on the morphological trait data collected in 2019-

2022. We identified and summarized the significant SNPs detected for each trait, and proposed a 

refined list of candidate genes that could be useful for future research. In many cases, QTL 

identified from this study overlapped with prior QTL mapping studies, providing confidence in 

the analyses and further support for QTL and candidate genes in the literature. We also identified 

potential novel genes important for these traits. The results from this work provide a broad view 

of the diversity of fruit quality traits in the core collection. The SNPs identified and the candidate 

genes proposed can be further functionally tested for more in depth understanding of the genetic 

mechanisms of these traits. 

124