Clemson University TigerPrints

All Theses Theses

August 2020

Enzymatic Targets for Improvement of

Andrew Disharoon Clemson University, [email protected]

Follow this and additional works at: https://tigerprints.clemson.edu/all_theses

Recommended Citation Disharoon, Andrew, "Enzymatic Targets for Improvement of Malting Sorghum" (2020). All Theses. 3376. https://tigerprints.clemson.edu/all_theses/3376

This Thesis is brought to you for free and open access by the Theses at TigerPrints. It has been accepted for inclusion in All Theses by an authorized administrator of TigerPrints. For more information, please contact [email protected].

ENZYMATIC TARGETS FOR IMPROVEMENT OF MALTING SORGHUM —————————————————————————————— A Thesis Presented to the Graduate School of Clemson University —————————————————————————————— In Partial Fulfillment of the Requirements for the Degree Master of Science Plant and Environmental Sciences —————————————————————————————— by Andrew Oakes Disharoon August 2020 —————————————————————————————— Accepted by: Dr. Stephen Kresovich, Committee Chair Dr. Richard Boyles Dr. William Bridges Dr. Susan Duckett

ABSTRACT

Sorghum bicolor (L.) Moench is a climate resilient originating from that has traditionally been used to make a variety of fermented beverages such as pito and . In western markets, the use of sorghum to produce and beverages has recently risen in part due to the visibility of a sensitive/intolerant market and a growing interest in unique inputs for beverage production. As such, there is a developing body of research on sorghum as a malted input into beverages. A major limitation to the wider adoption of sorghum as a substrate in is its low activity of amylolytic enzymes, either the result of insufficient content or inhibition by endogenous compounds. A collection of 42 diverse accessions representing the grain sorghum diversity panel, was evaluated for associations between alpha and beta amylase content, race, origin, and color as well as two classes of amylase inhibitors, phenols and . Among these accessions are several commonly used genetic resources, including reference line BTx623 and parents from the sorghum grain nested association panel.

Notable findings include accessions with high alpha amylase content, sources that may harbor additional high-amylase , associations with grain color, and populations which may be used to genetically map loci associated with the amylase activity.

These directions include the use of the sorghum bioenergy nested association panel to genetically map the trait and in-vivo methods to evaluate putative casual genetic loci.

ii ii DEDICATION

I would like to dedicate this work to my family, friends, and loved ones who supported me along this journey with its many unexpected twists and turns. And to , without

which this not only wouldn’t have been possible, it wouldn't have been much fun.

iii iii ACKNOWLEDGMENTS

I would like to acknowledge Clemson University’s Genetics and Biochemistry and Plant and Environmental Science departments for their assistantship support. I would also like to acknowledge the generous support provided by the Wade Stackhouse

Fellowship and those who have made this fellowship possible. I would also like to acknowledge all the mentors, collaborators, and co-workers who have aided in this work.

Finally, I would like to deeply thank all of those that have helped revise and proof this work.

iv iv TABLE OF CONTENTS

Page

TITLE PAGE ...... i

ABSTRACT...... ii

DEDICATION ...... iii

ACKNOWLEDGMENTS ...... iv

LIST OF TABLES ...... vii

LIST OF FIGURES ...... viii

CHAPTER

I. LITERATURE REVIEW ...... 1

Sorghum as a crop ...... 1 Sorghum germplasm and diversity ...... 2 Traditional uses of sorghum in fermented beverage making ...... 3 Why use sorghum to make beer?...... 5 The malting process ...... 7 Amylases ...... 11 Previous sorghum malting studies ...... 15 Summary and objectives ...... 16 References ...... 17

II. EXPLORING DIVERSE SORGHUM [ (L.) MOENCH] ACCESSIONS FOR AMYLASE ACTIVITY ...... 27

Introduction ...... 27 Material and methods...... 31 Results and discussion ...... 34 Conclusions ...... 42 References ...... 43

III. FUTURE DIRECTIONS ...... 51

An alternative to improve beta amylase in sorghum ...... 51 Potential, evaluation, and incorporation of malting germplasm ...... 52

v v Table of Contents (Continued) Page

Linkage mapping with recombinant inbred lines ...... 54 Genetic mapping with nested association mapping populations ...... 56 Causal gene identification ...... 59 Alternatives for a potentially complex trait ...... 64 Summary and conclusions ...... 66 References ...... 67

APPENDICES ...... 75

A: Figures...... 76 B: Tables ...... 81

vi vi LIST OF FIGURES

Figure Page

A-1 Mean amylase levels observed in materials samples as compared to plant introgression number ...... 76

A-2 Geographic distribution of materials sampled originating in Africa and their classification into regions ...... 77

A-3 Statistical difference between amylase content and malts of different colors ...... 78

A-4 Phenolic concentration and concentration compared to amylase content...... 79

A-5 Geographic distribution of material surveyed in this study ...... 80

vii vii LIST OF TABLES

Table Page

B-1 USDA sorghum beer germplasm ...... 81

B-2 Amylase study information ...... 85

B-3 Amylase content of bioenergy NAM parents ...... 86

viii viii CHAPTER ONE

LITERATURE REVIEW

Sorghum as a crop

Sorghum is a C4 cereal crop that is gluten free and belongs to the family Poaceae.

It is commonly known for its climate resilience and high photosynthetic efficiency

(Paterson et al., 2009). Through evolutionary diversification and plant breeding practices sorghum has been adapted to a broad range of cultivation conditions, including sub- tropical and temperate environments (Morris et al., 2013). With increasing global temperature and changing climate, the innate ability of the species to tolerate stressful abiotic conditions is likely to drive increased cultivation.

Currently, over half a billion people rely on sorghum to obtain a majority of their daily caloric intake. Sorghum grain is composed of approximately from 8.1 to 18.8% , 1.0 to 4.3% , and 61.7 to 71.1% , with the remainder being ash (Rhodes et al., 2017). This distribution of macromolecule content is comparable to that of other tropical such as or (Taylor et al., 2013). However, there is a wide array of variation in these compositional components observed among diverse sorghum accessions. There is also a notable diversity of secondary metabolites in the species

(Zhou et al., 2020), such as various phenolic compounds, including tannins, that have several affects to human health, nutrition quality, and biotic defense (Rhodes et al, 2014).

Sorghum accounts for 43% of all grain produced in Africa, being a dominant staple there

(Iqbal & Iqbal, 2015), and is the fifth most important cereal crop worldwide. It is

1 cultivated for a wide variety of uses including human consumption, production of biofuels, building/fiber materials, animal feed, as well as alcoholic beverage production

(Paterson et al., 2009). Other than being a commodity, sorghum is also an important model organism for other crop species in part due to its small genome size of approximately 750 Mb and low genomic complexity (Paterson et al., 2009). This simplicity has aided scientific efforts in the evaluation of many traits and has allowed for the foundations of numerous genetic studies.

Sorghum germplasm and diversity

Most of the widely used sorghum accessions for breeding in the United States

(US) come from the sorghum conversion program (SCP). The SCP is a collection of exotic sorghum lines in which dwarfing genes and photoperiod insensitivity have been introgressed through backcrossing and later through use of marker assisted selection

(Klein et al., 2016; Rosenow et al., 1997).

Photoperiod insensitivity allows the material to grow in the temperate climate of the US and short stature makes it amenable to combined harvesting. Non-converted photoperiod sensitive material can reach heights exceeding fifteen feet tall which makes them cumbersome to work with, and they will not flower in the United States due to growing degree day requirements (Klein et al, 2008). The SCP was initiated with a goal to provide genetic diversity to breeding programs and to increase the productivity of grain sorghum hybrids (Klein et al., 2016; Rosenow et al., 1997).

2 Any phenotype collected from such a diverse set of accessions serves as a means of comparison to that of other sorghum germplasm. The SCP accessions are one of the primary resources for studying and obtaining exotic germplasm for temperate sorghum breeding programs. Therefore, many of these accessions have been incorporated into genetic mapping resources. These include the bioenergy mapping association population

(unpublished), the sorghum association panel (SAP) (Casa et al., 2008), and the SAP expansion into the grain sorghum diversity panel (GSDP) (Boyles et al., 2017). We pulled material from these carefully constructed populations and preserved the selection criteria used to make them so robust conclusions can be made. This allows future studies to build from this work and, hopefully, discover the genetic underpinnings of amylase variation by expanding the evaluation to the larger diversity panels. Identifying genetic markers associated with amylase variation will allow breeders to optimize their efforts through marker assisted selection.

Traditional uses of sorghum in fermented beverage making

“What is beer?” This question is a topic of endless debate, despite regulations including the German Reinheitsgebot, which have attempted to define the beverage

(Gaab, 2006). Here, I reference to two classes of fermented beverages, defined imperfectly due to their immense and wonderful diversity. “Traditional beverages,” and variations, are defined as fermented beverages that are generally outside the global or western palette, with some notable exceptions. “Western- or global-styled and clarified beverages” encompass the remainder. Sorghum is used to make a variety of beverages:

3 traditional beverages including baijiu, , and pito, and to a lesser degree, western-styled beers such as Anheuser-Busch’s Redbridge™.

Looking at traditional beverages, baijiu is a distilled liquor commonly made from red sorghums in dating back to the Yuan Dynasty (Schmalzer, 2002). As of 2018, this liquor is produced upwards of 10.8 billion liters, which exceeds total production of vodka, whiskey, rum, tequila and gin combined (Zheng & Han, 2016; The Jakarta Post, n.d.). Other fermented sorghum beverages, such as the traditional beers consumed across

Africa, have equally long histories of production (Rogerson, 2019). Even in modern production throughout Africa, typical western style inputs such as and are ill adapted to the region compared to sorghum. Thus, sorghum remains a significant portion of the substrate for in many beverages produced there. Furthermore, events such as the 1970’s embargo of barley in have increased the price of modern fermentation components to restrictive levels (Abuja, 2019).

Despite this, several of the world's largest commercial beer production corporations are now focused on Africa as a potential market for expansion. Their efforts have already endangered the existence of several traditional sorghum beers, including the

South African umqombothi (Slow foods, 2017). Regardless, numerous other traditional

African sorghum beverages are unlikely to be forgotten due to their number and strong cultural history. This long-standing use of sorghum to produce fermented beverages may have resulted in the selection of varieties which are excellent for that purpose. In fact, several sorghum accessions in the US National Plant Germplasm System have been

4 introduced from Africa due to their purported quality in beverage making. These are highlighted in table B-1.

Much of the traditional beer in Africa is made using a process which relies not only on the fermentation of yeasts, but also bacteria, which are able to degrade of sorghum (Lyumugabe et al., 2012; Rogerson, 2019). Many of these beverages are consumed in an active fermentation state, similar to a cultured product, and with unprocessed solids still in them. These products have not found favor outside of their traditional markets, perhaps due in part to their unique flavors and textures. However, there are some African sorghum beverages that do not employ bacteria, such as pito, a clarified sorghum beer more akin to western-styled beers.

These beverages are often made from sorghum with high diastatic power (a measure of saccharification potential), as reported by Djameh et al. 2015, which allows them to partially overcome the need for bacteria.

Why use sorghum to make beer?

While many use sorghum in fermented beverages due to its availability, there are additional reasons for its use in malting. With increasing pressures due to climate change on other malting , sorghum's resilience to heat and drought stresses might lead to its increased use in malt production and incorporation as an adjunct (Zhao et al., 2017).

Compared to barley, which is the most widely used grain to make beer, sorghum yields are less impacted by global temperature changes (Lobell & Field, 2007). Wheat and

5 barley also suffer high yield losses under water stress caused by drought, unlike sorghum which is tolerant to these conditions (De Souza et al., 2015; Hadebe et al., 2017).

In addition to production needs, consumer wants are also driving sorghum’s adoption. The species is host to a variety of compounds that result in a unique flavor profile when used in the production of foodstuffs (Brannan et al., 2001; Lasekan et al.,

1997). These compounds and flavors are one component influencing sorghum adoption by craft producers. While there is a relative lack of research on sorghum flavor contributions to fermented beverages, much of the current commercial sorghums used as malt in the US likely have low levels of positive organoleptic traits. However, the more diverse sorghums of Africa are prized for their flavor contributions in their traditional beverages (Lyumugabe et al., 2012; Smith & Frederiksen, 2000).

The rise in the use of sorghum for fermented beverages has also been driven by the gluten sensitive/intolerant market (360 Research Reports, n.d.). In sufficient quantities, both rice and corn adjuncts can elicit negative reactions from those who are gluten sensitive. Thus many brewers produce gluten free beers made with only sorghum, common , or (Arendt & Dal Bello, 2011). With approximately 3 million people in the US alone being gluten intolerant, and 18 million considered gluten sensitive

(Mansueto et al., 2014; Presutti et al., 2007), improving sorghum for the purpose of beverage production directly impacts these various markets for both flavor and health.

However, sorghum has several limitations that hinder its acceptance for more broad use in beverage production. First, several compounds within sorghum are undesirable for both taste and production. Tannins, which are rife within many sorghums,

6 are both anti-nutritive to yeasts and are astringent and bitter in taste (Cheng et al., 2009;

Dykes et al., 2005). In certain products these organoleptic qualities are desirable (Cheng et al., 2009; Oliver & Colicchio, 2011; Smith & Frederiksen, 2000), however in the vast majority of beverages the tannins’ contribution to taste is unacceptable. Other compounds, including phenolics, contribute “band-aid like” and medicinal flavoring to the beverage and are undesirable (Soares et al., 2013). However, sorghums with low tannin and phenolic content are present within the germplasm available to breeders and are traits present in elite production . Beyond negative flavor, one of the largest challenges for using sorghum in beverage production is the low levels of saccharifying enzymes in the malts of commercial sorghums (Beta et al., 1995; Dufour et al., 1992).

This low-level slows yeast growth as insufficient soluble are released in the wort, which is the cooked slurry of malt, water, hops, and any other additives. These challenges of flavor and enzymatic activity are possible to overcome with tools and knowledge currently available to sorghum breeders.

The malting process

The process of malting causes significant changes in composition and metabolism to grains as they prepare for germination. These changes are necessary to make nutrients stored within the grain available to microorganisms for fermentation, just as they would naturally make nutrients available for the developing embryo (Kunze, n.d.; Oliver &

Colicchio, 2011). In the process, proteins, , , and micronutrients are all modified to be more accessible. Among the most important of these changes is the

7 breakdown of complex storage molecules into simpler ones that can be readily used by the developing embryo. For malting, the most important among these carbohydrates are starches, which act as the primary source of soluble sugars (Narziss &

Back, 2019). While making the fermented beverage, these sugars fuel yeast activity in the wort.

The process of malting begins with grain that is largely homogenous: free from disease, debris, and damage. Depending on the species being malted, there are also preferences for grain macronutrient composition. For example, barley malts generally must fall within 10% and 12% protein to be accepted by malthouses (Osman, 2003).

Protein content outside this range is either insufficient to feed yeast or results in hazy beer and off flavors (Steiner et al., 2011). Grains are then hydrated in a process often called steeping. The length of hydration may vary between species, but generally lasts several hours to a day (Oliver & Colicchio, 2011), but can be over a longer period as is seen with sorghum malting studies (Dewar et al., 1997). Steeping activates the germination process in the seed and provides the necessary hydration. Temperature is often controlled during this process to maximize the germination of the grains, as it too is a factor that regulates grain germination in many crops (Agu & Palmer, 1997; Dewar et al., 1997). Sorghums are steeped at a wide range of temperatures commercially, but studies have shown that greater production of saccharifying enzymes occurs at 25° to 30°

C and increase with longer steeping times (Dewar et al., 1997). Furthermore, steeping with a solution at pH 6.5 has also been shown to increase the activity of one such enzyme, alpha amylase, in both green and kilned sorghum malts (Adefila et al., 2012).

8 Steeping solutions may also include additives, such past use of dilute formaldehyde, to control for the development of microorganisms or to deactivate undesired compounds in the grain (Beta et al., 1995; Lyumugabe et al., 2012).

After steeping, grains are moved to a germination chamber where they undergo the majority of their compositional changes. Germination periods vary between grain species, but literature suggests that a five day germination period is sufficient for sorghum (Beta et al., 1995). Temperatures are also typically controlled during this period, with studies showing sorghum responds well to germination temperatures between 20° to

30° C (Adefila et al., 2012; Aisen & Muts, 1987; Beta et al., 1995). The process of germination causes a large metabolic shift in the grain, including the marked production of enzymes that drive this process. For example, in sorghum, prior to malting there is almost no activity of beta amylase, whereas after malting it appears in detectable concentrations (Dufour et al., 1992; J. R. N. Taylor & Robbins, 1993). The germination process also results in the emergence of seedling hypocotyls and radicals which are later removed. As part of the plant’s metabolism, some of the stored energy of the seed is lost due to respiration. Both the removal of the emerged plant parts and respiration ultimately result in loss of energy that can no longer be used by the yeast (Narziss & Back, 2019).

Therefore, it is desirable to keep germination as short as possible while also providing enough time to release maximal saccharifying enzymes.

After germination, grains are then dried, or kilned, in an oven. This reduces the water content of the seed, halts germination, inactivates certain biological actors including many enzymes, and drives off many volatile compounds that are considered

9 unpalatable (Oliver & Colicchio, 2011). This process also prevents spoilage of the grains, which are quite susceptible to attack by microorganisms and fungi now that they are replete with easily digestible nutrients.

Kilning can take from 20 hours to two days and occurs at a range of temperatures that are selected based on the final product qualities desired (Oliver & Colicchio, 2011).

Higher temperatures and longer times result in more reactions between sugars and proteins in the malts, a class of chemical reactions known as Maillard reactions. This produces many favorable flavors, including the rich taste of dark roast and seared steak (van Boekel, 2006). In malts, these reactions also provide the flavors which distinguish many differences between beer varieties, as darker beers experience more

Maillard reactions (Coghe et al., 2004). As the length and temperature of kilning greatly determine the taste of the malt, there is a great deal of variation in this parameter of the malting process as well.

There are innumerable variations to this process in time, temperature, and additives. Furthermore, for some products malting involves additional steps to add unique characteristics to the final product, such as with the production of caramel malts. These variations all impact the metabolomic activity of enzymes in the grain and ultimately change the final malt flavor, nutrition, and ease to work with. The exact species used, how it was grown, and many other factors also account for the final quality of the malt

(Bettenhausen et al., 2018). Ultimately, the malting process and the resulting best malts are the ones that are cost efficient, flavorful, and fit the needs of the brewer.

10 Amylases

As discussed above, there are many factors that relate to grain quality for malting.

However, the malt’s enzymatic capacity is one of the most important factors (Oliver &

Colicchio, 2011) in evaluating enzymes that modify the nutritional storage molecules in grain, amylases are the most important as they release sugars that are the lifeblood of yeast. There are two primary amylases that carry out this process of liquefaction and saccharification: alpha and beta amylase (Oliver & Colicchio, 2011). Alpha amylase breaks down large carbohydrates, the starches, into smaller ones such as dextrins, solubilizing them. They do this by the action of random hydrolysis of starch chains at alpha (1,4) glucosidic bonds distanced from chain ends.

However they cannot hydrolyze alpha (1,6) linked branches off from the main starch chain (Hough et al., 1982) rife in some sorghums. These smaller sugars which are released by alpha amylases activity are still less than ideal for yeast metabolism and must be further broken down by the activity of beta amylases. Beta amylases hydrolyzes the penultimate alpha (1,4) glucosidic bond from the non-reducing end of the sugars released by alpha amylase, releasing beta-maltose. Starches are first broken first into soluble sugars by alpha amylase, that are easily acted upon by other enzymes now that they are soluble in liquid mash. Beta amylase then breaks down polysaccharides released by fermentable sugars (disaccharides) by beta amylase can be used by the microorganisms during fermentation. The action of beta amylase and other enzymes spawn these fermentable sugars which include glucose, fructose, sucrose, maltose, and maltotriose.

Such saccharides account for approximately 65% of the total dissolved solids in the wort.

11 In discussing these enzymes, it is also important to note the substrate that they act on, starch, and the major forms it takes. Sorghums with greater content of amylopectin

(starch molecules with numerous branches and cross linkages) are considered waxy, while those with fewer branches (having greater amylose content), are considered floury.

Sorghum total grain composition of amylose averages 17%, while amylopectin averages around 50%, however these show appreciable variation among accessions (Dicko et al.,

2005). These difference in starch composition have impacts on how these starches behave when cooked, how enzymes can act upon the starch molecules, how readily digested the carbohydrates are (Mezgebe et al., 2018).

Specifically notable for sorghum , greater amounts of amylopectin appear to bring the optimal activity temperatures of alpha and beta amylase in mash together by approximately 2 °C due to grain swelling (Taylor et al., 2013), lessening an issue with sorghum malt-based mashes with incongruent maximum activation temperatures of alpha amylase and the inactivation temperature of beta amylase. In detraction from the use of abundantly waxy sorghums, both alpha and beta amylase cannot hydrolyze the branching bonds of starch (alpha-[1,6] linkages) and therefore in isolation, result in incomplete hydrolysis of amylopectin (Dicko et al., 2005) at higher rates in sorghums with more amylopectin. Despite this, the loss is overall negligible in comparison to the overall hydrolysis of the remaining alpha (1,4) glucosidic bonds of starch into fermentable sugars. Thus, the discussion on alpha and beta amylase here is well deserved as their activities bear the brunt of the burden for fermentation success.

12 Sorghum beta amylases are found in relatively low concentrations in malts as observed by Beta et al. (1995) and others. Further complicating the story of sorghum beta amylase, the temperature at which sorghum alpha amylase is most active, 70° C, is just above the deactivation temperature of sorghum beta amylase at 68° C (El Nour &

Yagoub, 2010; Kumar et al., 2005).

Thus, in achieving complete function of sorghum alpha amylase on wort, sorghum beta amylase will no longer be viable (John R. N. Taylor et al., 2013). Despite this, we endeavored here to identify sorghums that exhibit variation for both amylases that breeders can capitalize on for the improvement of the traits. Other enzymes such as beta-glucosidases, which are responsible for the degradation of cell walls, fall in the same category of being insufficiently produced by sorghum for western-style beverages

(Etokakpan & Palmer, 1990).

Although yeasts with significant amylolytic ability for both dextrin and starches have been isolated, (Laluce et al., 1988), the commonly used brewer’s yeast,

Saccharomyces cerevisiae, cannot normally degrade these molecules. To compensate, many traditional African beers use other microorganisms to provide the additional necessary saccharification activity that sorghum does not provide (Lyumugabe et al.,

2012; Rogerson, 2019). Many of such organisms are used in the production of varieties of global/western beers. For example, Lactobacillus are able to degrade starches through the production of secreted enzymes, aiding in the production of beverages such as Flanders’

Red Ale™ (Gänzle & Follador, 2012). The lack of amylase activity in sorghum malts has been circumvented by the recruitment of these organisms, however, this method is

13 insufficient for global production needs nor for all styles of beer. These organisms produce their own unique metabolites that contribute to flavors to beer and are not acceptable in all beer varieties. Therefore, there is still a substantial need to increase endogenous amylase activity in sorghum malts.

If the yeasts most broadly used in industry cannot optimally use the sugars produced by sorghum’s alpha amylase alone, why evaluate it? Alpha amylase still holds significance to malting as a means of providing available carbohydrates. The more completely the starches are broken down to sugars such as dextrins, the more available carbohydrates there are to be used. Even if supplemental beta amylase is required, the comparably low levels of alpha amylase in most sorghum accessions preclude sufficient beta amylase function. Without alpha amylase activity, beta amylases will not have targets to degrade at all. In most sorghums, starch degradation cannot be accomplished without adding supplementing enzymes or an adjunct grain with sufficient alpha amylase activity. The improvement of food grade sorghums with high alpha amylase genes from diverse sorghum cultivars would reduce the amount of adjunct grain or supplemental enzymes needed, and thus lower the cost to the brewer.

While both alpha and beta amylases have significance in beverage production, sorghum alpha amylases in particular have been implicated for other applications.

Sorghum alpha amylases are uniquely thermostable and therefore can be used in a variety of industrial applications and bioenergy production (Kumar et al., 2005). They can be isolated and employed in the digestion of organic biomass for bioenergy pipelines and have been described as a credible substitute for microbially derived amylases in this

14 application (Adefila et al., 2012). The variety of uses for sorghum’s alpha amylase and its importance in malting highlight the need to study the enzyme over less impactful ones such as beta amylase.

Previous sorghum malting studies

Sorghum has been studied extensively for its use as a malt and substrate for brewing, with many well cited for publications on the topic (Adefila et al., 2012; Aisen &

Muts, 1987; Beta et al., 1995; Dewar et al., 1997; Dufor & Srebrnik 1992; Ogu et al.

2006). Additionally, several historical events, including the embargo of imported barley in Nigeria in the 1970s, have created renewed interest in the topic at regular intervals.

However, quantity does not always indicate quality, and many commonly cited studies have not adequately evaluated the diversity of sorghum. Thus, they have obtained results which are not representative of the sorghum's true potential.

Among the well cited studies on sorghum malting is a paper by Beta et al. 1995 that used a collection of materials composed of 16 cultivars grown in the Texas

Agricultural Station in Lubbock TX in 1992. While the materials in this study were grown under identical conditions, the diversity of the cultivars was limited, restricted mostly to breeding materials commonly used in the United States. However, this publication is widely cited as providing an expected range of alpha amylase activity in sorghum. Another key study by Dufor & Srebrnik in 1992, which used a greater diversity of material, evaluated more than 49 sorghum cultivars. However, these samples largely came largely from two regions: and Rwanda. Additionally, these samples were

15 not grown in the same locality, and as reported by the authors this environmental difference likely contributed much of the variation between samples. One more recent study by Ogu et al. in 2006 evaluated a collection of only four sorghums for alpha amylase activity, some of which had been previously studied in Beta et al. 1995, and provided little to our understanding of the trait within a diverse population. There are many other studies which examine the optimization of the malting process itself (Adefila et al., 2012; Aisen & Muts, 1987; Dewar et al., 1997), however, these also survey little variety within sorghums. Because these and several other studies underrepresented sorghum diversity, I wished to study an improved collection based on diverse origins, racial classifications, and use history, that preserved the diversity used to construct the

GSDP.

Summary and objectives

Sorghum is a diverse crop with respect to its genetics and end-use products. One of its oldest applications as a substrate for malted beverages is derived from a long tradition and has a wide international cultural impact. In light of the continued climate change expected in the coming decades, the significant economic impact of sorghum based products will likely continue to grow due to its resilience for such expected stresses

(Ceccarelli et al., 2010). Similarly, demands based on specific consumer needs and a desire for more diverse inputs into products is also driving the adoption of malting sorghum. However, sorghum has several challenges in its malting quality compared to other grains. While sorghum may never be fully malted at a capacity that is comparable

16 to barley, it does show significant variation in one of the most important traits for malt quality, suggesting that alpha amylase activity can be optimized. Alpha amylase in sorghums shows the potential to not only improve malting, but to also improve bioenergy pipelines. As such, the improvement of this characteristic is an immediately tangible goal with wide reaching implications. While other studies have evaluated alpha amylase activity in sorghum, many of these studies fail to evaluate samples that are representative of the breath of sorghum diversity. In this study, I used a subset of accessions from the sorghum diversity panel (SAP/GSDP) which maintains the intended genetic diversity of the original population to examine and expand upon previous sorghum malting studies. In evaluating a more diverse collection of material, I have identified potential populations that harbor variation in alpha amylase activity and can be harnessed by breeders to improve enzyme productivity. Looking forward, the use of material from the GSDP allows other phenotypes collected on these accessions to be compared to those collected in this work.

References

Abuja, N. I. (2019, November 5). Ban on importation of barley malt saves Nigeria $500m

annually — Onu. Leadership Newspaper. https://leadership.ng/2019/11/05/ban-

on- importation-of-barley-malt-saves-nigeria-500m-annually-onu/

Adefila, O. A., Bakare, M. K., & Adewale, I. O. (2012). Characterization of an α-amylase

from sorghum (Sorghum bicolor) obtained under optimized conditions. Journal of

the Institute of Brewing, 118(1), 63–69.

17 Agu, R. C., & Palmer, G. H. (1997). The effect of temperature on the modification of

sorghum and barley during malting. Process Biochemistry, 32(6), 501–507.

https://doi.org/10.1016/S0032-9592(97)00002-2

Aisen, A. O., & Muts, G. C. J. (1987). Micro-scale malting and brewing studies of some

sorghum varieties. Journal of the Institute of Brewing, 93(4), 328–331.

https://doi.org/10.1002/j.2050-0416.1987.tb04513.x

Arendt, E., & Dal Bello, F. (2011). Gluten-free cereal products and beverages. Elsevier.

Beta, T., Rooney, L. W., & Waniska, R. D. (1995). Malting characteristics of sorghum

cultivars. Cereal Chemistry, 72(6), 533–538.

Bettenhausen, H. M., Barr, L., Broeckling, C. D., Chaparro, J. M., Holbrook, C., Sedin,

D., & Heuberger, A. L. (2018). Influence of malt source on beer chemistry,

flavor, and flavor stability. Food Research International, 113, 487–504.

https://doi.org/10.1016/j.foodres.2018.07.024

Boyles, R. E., Pfeiffer, B. K., Cooper, E. A., Rauh, B. L., Zielinski, K. J., Myers, M. T.,

Brenton, Z., Rooney, W. L., & Kresovich, S. (2017). Genetic dissection of

sorghum grain quality traits using diverse and segregating populations.

Theoretical and Applied Genetics, 130(4), 697–716.

Brannan, G. L., Setser, C. S., Kemp, K. E., Seib, P. A., & Roozeboom, K. (2001).

Sensory characteristics of grain sorghum hybrids with potential for use in human

food. Cereal Chemistry, 78(6), 693–700.

https://doi.org/10.1094/CCHEM.2001.78.6.693

18 Brenton, Z. W., Cooper, E. A., Myers, M. T., Boyles, R. E., Shakoor, N., Zielinski, K. J.,

Rauh, L., Bridges, W. C., Morris, G. P., & Kresovich, S. (2016). A genomic

resource for the development, improvement, and exploitation of sorghum for

bioenergy. Genetics, 204(1), 21–33. https://doi.org/10.1534/genetics.115.183947

Casa, A. M., Pressoir, G., Brown, P. J., Mitchell, S. E., Rooney, W. L., Tuinstra, M. R.,

Franks, D., & Kresovich, S. (2008). Community resources and strategies for

association mapping in sorghum. Crop Science, 48(1), 30–40.

https://doi.org/10.2135/cropsci2007.02.0080

Ceccarelli, S., Grando, S., Maatougui, M., Michael, M., Slash, M., Haghparast, R.,

Rahmanian, M., Taheri, A., Al-Yassin, A., Benbelkacem, A., Labdi, M., Mimoun,

H., & Nachit, M. (2010). Plant breeding and climate changes. The Journal of

Agricultural Science, 148(6), 627–637.

https://doi.org/10.1017/S0021859610000651

Slow Foods. Centuries-old and steeped in tradition, Umqombothi joins the Ark. (2017,

February 12). Slow Food Foundation. https://www.fondazioneslowfood.com/en/

centuries-old-and-steeped-in-tradition-umqombothi-joins-the-ark/

Cheng, S., Sun, Y., & Halgreen, L. (2009). The relationships of sorghum kernel pericarp

and testa characteristics with tannin content. Asian J. Crop Sci, 1(1), 1–5.

Coghe, S., Derdelinckx, G., & Delvaux, F. R. (2004). Effect of nonenzymatic browning

on flavour, color and antioxidative activity of dark specialty malts — a review.

Monatsschrift fur Brauwissenschaft. 57. 25–38.

19 De Souza, A. P., Cocuron, J.-C., Garcia, A. C., Alonso, A. P., & Buckeridge, M. S.

(2015). Changes in whole-plant metabolism during the grain-filling stage in

sorghum grown under elevated CO2 and drought [OPEN]. Plant Physiology,

169(3), 1755–1765. https://doi.org/10.1104/pp.15.01054

Dewar, J., Taylor, J. R. N., & Berjak, P. (1997). Determination of improved steeping

conditions for sorghum malting. Journal of Cereal Science, 26(1), 129–136.

https://doi.org/10.1006/jcrs.1996.0101

Dufour, J. P., Melotte, L., & Srebrnik, S. (1992). Sorghum malts for the production of a

lager beer. Journal of the American Society of Brewing Chemists, 50(3), 110–119.

Dykes, L., Rooney, L. W., Waniska, R. D., & Rooney, W. L. (2005). Phenolic

compounds and antioxidant activity of sorghum grains of varying genotypes.

Journal of Agricultural and Food Chemistry, 53(17), 6813–6818.

https://doi.org/10.1021/jf050419e

El Nour, M. E., & Yagoub, S. O. (2010). Partial purification and characterization of α-

and β- amylases isolated from Sorghum bicolor cv.(Feterita) malt. Journal of

Applied Sciences, 10(13), 1314–1319.

Etokakpan, O. U., & Palmer, G. H. (1990). Comparative studies of the development of

endosperm-degrading enzymes in malting sorghum and barley. World Journal of

Microbiology and Biotechnology, 6(4), 408–417.

https://doi.org/10.1007/BF01202124

Gaab, J. S. (2006). Munich: Hofbräuhaus & History: Beer, Culture, & Politics. Peter

Lang.

20 Gänzle, M. G., & Follador, R. (2012). Metabolism of oligosaccharides and starch in

lactobacilli: A review. Frontiers in Microbiology, 3.

https://doi.org/10.3389/fmicb.2012.00340

360 Research Reports. (n.d.). Gluten-Free Beer Market | Size | Trend | Growth to 2023.

Retrieved January 29, 2019, from https://www.mordorintelligence.com/industry-

reports/gluten-free-beer-market

Hadebe, S. T., Modi, A. T., & Mabhaudhi, T. (2017). Drought tolerance and water use of

cereal crops: A focus on sorghum as a food security crop in sub-Saharan Africa.

Journal of Agronomy and Crop Science, 203(3), 177–191.

https://doi.org/10.1111/jac.12191

Hough, J. S., Briggs, D. E., Stevens, R., & Young, T. W. (1982). Malting and Brewing

Science, Volume II: Hopped Wort and Beer. Springer US.

https://doi.org/10.1007/978-1-4615- 1799-3

Iqbal, M. A., & Iqbal, A. (2015). Overview on sorghum for food, feed, forage and :

Opportunities and problems in ’s perspectives. Am. Eurasian J. Agric.

Environ. Sci, 15, 1818–1826.

Klein, R. R., Miller, F. R., Bean, S., & Klein, P. E. (2016). Registration of 40 converted

germplasm sources from the reinstated sorghum conversion program. Journal of

Plant Registrations, 10(1), 57–61. https://doi.org/10.3198/jpr2015.05.0034crg

Kumar, R. S. S., Singh, S. A., & Rao, A. G. A. (2005). Thermal stability of α-amylase

from malted jowar (Sorghum bicolor). Journal of Agricultural and Food

Chemistry, 53(17), 6883–6888. https://doi.org/10.1021/jf0501701

21 Klein, R. R., Mullet, J. E., , D. R., Miller, F. R., Rooney, W. L., Menz, M. A.,

...Klein, P. E. (2008). The effect of tropical sorghum conversion and inbred

development on genome diversity as revealed by high-resolution genotyping.

Crop Sci., 48(S1), S-12– S-26. doi: 10.2135/cropsci2007.06.0319tpg

Kunze, W. (n.d.). Brewing Malting.

Laluce, C., Bertolini, M. C., Ernandes, J. R., Martini, A. V., & Martini, A. (1988). New

amylolytic yeast strains for starch and dextrin fermentation. Applied and

Environmental Microbiology, 54(10), 2447–2451.

Lasekan, O. O., Lasekan, W. O., & Idowu, M. A. (1997). Flavour volatiles of ‘malt

beverage’ from roasted sorghum. Food Chemistry, 58(4), 341–344.

https://doi.org/10.1016/S0308- 8146(96)00202-6

Lobell, D. B., & Field, C. B. (2007). Global scale climate–crop yield relationships and

the impacts of recent warming. Environmental Research Letters, 2(1), 014002.

Lyumugabe, F., Gros, J., Nzungize, J., Bajyana, E., & Thonart, P. (2012). Characteristics

of African traditional beers brewed with sorghum malt: A review. Biotechnologie,

Agronomie, Société et Environnement, 16(4).

https://orbi.uliege.be/handle/2268/142889

Mansueto, P., Seidita, A., D'Alcamo, A., & Carroccio, A. (2014). Non-celiac gluten

sensitivity: literature review. J. Am. Coll. Nutr., J(Am), Coll.

https://10.1080/07315724.2014.869996

22 Mezgebe, A. G., Abegaz, K., & Taylor, J. R. N. (2018). Relationship between waxy (high

amylopectin) and high protein digestibility traits in sorghum and malting quality.

J. Cereal Sci., 79, 319–327. https://10.1016/j.jcs.2017.11.015

Morris, G. P., Ramu, P., Deshpande, S. P., Hash, C. T., Shah, T., Upadhyaya, H. D.,

Riera- Lizarazu, O., Brown, P. J., Acharya, C. B., Mitchell, S. E., Harriman, J.,

Glaubitz, J. C., Buckler, E. S., & Kresovich, S. (2013). Population genomic and

genome-wide association studies of agroclimatic traits in sorghum. Proceedings

of the National Academy of Sciences, 110(2), 453–458.

https://doi.org/10.1073/pnas.1215985110

Narziss, L., & Back, W. (2019). Die Bierbrauerei: Band 1 - Die Technologie der

Malzbereitung. John Wiley & Sons.

Oliver, G., & Colicchio, T. (2011). The Oxford companion to beer. Oxford University

Press.

Osman, A. M. (2003). Barley and malt proteins and proteinases: III. A simple method for

estimating the combined actions of malt proteinases and the extent of protein

degradation during malting. Journal of the Institute of Brewing, 109(2), 150–153.

https://doi.org/10.1002/j.2050-0416.2003.tb00145.x

Paterson, A. H., Bowers, J. E., Bruggmann, R., Dubchak, I., Grimwood, J., Gundlach, H.,

Haberer, G., Hellsten, U., Mitros, T., Poliakov, A., Schmutz, J., Spannagl, M.,

Tang, H., Wang, X., Wicker, T., Bharti, A. K., Chapman, J., Feltus, F. A., Gowik,

U.,… Rokhsar, D. S. (2009). The Sorghum bicolor genome and the diversification

of grasses. Nature, 457(7229), 551–556. https://doi.org/10.1038/nature07723

23 The Jakarta Post (n.d.). Tough sell: Baijiu, China’s potent tipple, looks abroad. Retrieved

March 6, 2020, from https://www.thejakartapost.com/life/2019/06/17/tough-sell-

baijiu-chinas-potent-tipple-looks-abroad.html

Presutti, R. J., Cangemi, J. R., Cassidy, H. D., & Hill, D. A. (2007). Celiac disease.

American Family Physician, 76(12), 1795–1802.

Rhodes, D. H., Hoffmann, Jr., L., Rooney, W. L., Ramu, P., Morris, G. P., Kresovich, S.,

...Kresovich, S. (2014). Genome-wide association study of grain polyphenol

concentrations in global sorghum [Sorghum bicolor (L.) Moench] Germplasm. J.

Agric. Food. Chem., 62(45), 10916–10927. doi: 10.1021/jf503651t

Rhodes, D. H., Hoffmann, L., Rooney, W. L., Herald, T. J., Bean, S., Boyles, R.,

...Kresovich, S. (2017). Genetic architecture of kernel composition in global

sorghum germplasm. BMC Genomics, 18(1), 1–8. doi: 10.1186/s12864-016-

3403-x

Rogerson, C. M. (2019). African traditional beer: Changing organization and spaces of

South Africa’s sorghum beer industry. African Geographical Review, 38(3), 253–

267. https://doi.org/10.1080/19376812.2019.1589735

Rosenow, D. T., Dahlberg, J. A., Stephens, J. C., Miller, F. R., Barnes, D. K., Peterson,

G. C., Johnson, J. W., & Schertz, K. F. (1997). Registration of 63 converted

sorghum germplasm lines from the sorghum conversion program. Crop Sci, 37(4),

1399–1400.

24 Schmalzer, S. (2002). Science and Civilisation in China. Vol. 6: Biology and Biological

Technology. Pt. 5: and Food Science (review). Technology and

Culture, 43(2), 407–409. https://doi.org/10.1353/tech.2002.0083

Smith, C. W., & Frederiksen, R. A. (2000). Sorghum: Origin, history, technology, and

production (Vol. 2). John Wiley & Sons.

Soares, S., Kohl, S., Thalmann, S., Mateus, N., Meyerhof, W., & De Freitas, V. (2013).

Different phenolic compounds activate distinct human bitter taste receptors.

Journal of Agricultural and Food Chemistry, 61(7), 1525–1533.

https://doi.org/10.1021/jf304198k

Steiner, E., Gastl, M., & Becker, T. (2011). Protein changes during malting and brewing

with focus on haze and foam formation: A review. European Food Research and

Technology, 232(2), 191–204. https://doi.org/10.1007/s00217-010-1412-6

Taylor, J. R. N., & Robbins, D. J. (1993). Factors influencing beta-amylase activity in

sorghum malt. Journal of the Institute of Brewing, 99(5), 413–416.

https://doi.org/10.1002/j.2050-0416.1993.tb01181.x

Taylor, John R. N., Dlamini, B. C., & Kruger, J. (2013). 125th Anniversary Review: The

science of the tropical sorghum, maize and rice in relation to lager beer

brewing. Journal of the Institute of Brewing, 119(1–2), 1–14.

https://doi.org/10.1002/jib.68 van Boekel, M. A. J. S. (2006). Formation of flavour compounds in the Maillard reaction.

Biotechnology Advances, 24(2), 230–233.

https://doi.org/10.1016/j.biotechadv.2005.11.004

25 Zhao, C., Liu, B., Piao, S., Wang, X., Lobell, D. B., Huang, Y., Huang, M., Yao, Y.,

Bassu, S., Ciais, P., Durand, J.-L., Elliott, J., Ewert, F., Janssens, I. A., Li, T., Lin,

E., Liu, Q., Martre, P., Müller, C.,… Asseng, S. (2017). Temperature increase

reduces global yields of major crops in four independent estimates. Proceedings

of the National Academy of Sciences, 114(35), 9326–9331.

https://doi.org/10.1073/pnas.1701762114

Zhou, Y., Wang, Z., Li, Y., Li, Z., Liu, H., & Zhou, W. (2020). Metabolite profiling of

sorghum of different colors from different cultivars using a

widely targeted metabolomics approach. Int. J. Genomics, 2020. doi:

10.1155/2020/6247429

Zheng, X.-W., & Han, B.-Z. (2016). Baijiu (白酒), Chinese liquor: History, classification

and manufacture. Journal of Ethnic Foods, 3(1), 19–25.

https://doi.org/10.1016/j.jef.2016.03.001

26 CHAPTER TWO

EXPLORING DIVERSE SORGHUM [Sorghum bicolor (L.) MOENCH]

ACCESSIONS FOR MALT AMYLASE ACTIVITY

Andrew Disharoona, Richard Boylesa, Kathleen Jordanb, Stephen Kresovich a# a: Department of Plant and Environmental Sciences, Clemson University, South Carolina, USA b: Advanced Plant Technology, Clemson University, Clemson, SC, USA

#: Corresponding/communicating author ([email protected])

Submitted to the Journal of the Institute of Brewing May 20, 2020

Introduction

Alpha and beta amylases are essential enzymes in the liquefaction and saccharification of wort, catalyzing hydrolysis of carbohydrates (starches) at the alpha-

1,4 glucosidic bonds internally and at penultimate sites respectively (1). Alpha amylase breaks down starches into smaller soluble sugars units that can be easily acted upon by beta amylase, which then further hydrolyses them into fermentable sugars (1). Combined, their activity is directly related to fermentation success and insufficient activity of these enzymes, either through their absence or inhibition by other compounds, results in low concentrations of bioavailable carbohydrates for fermenting organisms. Thus, insufficiency reduces the nutritional value of the malt in brewing and for derivative products such as syrups.

27 Sorghum bicolor (L.) Moench, is a climate resilient, starchy, gluten-free, cereal grain species that is used in many cultures to make malted beverages and foodstuffs, but the grain commonly displays poor self-sufficient saccharification in mashing (2,3,4). In the United States, there is a growing demand for unique beer inputs and gluten conscious products, which has in part driven increased demand for sorghum malts (5). However, most commercial sorghums have poor amylase activity necessary for brewing, especially compared to more common brewing substrates such as barley (2,4). While low amylase activity is one major inhibitor in sorghum wort’s fermentation success, their malts have other brewing challenges that limit their use as a brewing substrate, including high gelatinization temperature (6) and high mash viscosity (7).

Nevertheless, many cultures across the globe have used sorghum malt to produce traditional fermented beverages, such as jwala, pito, and baijiu (8,9). To overcome the challenges of sorghum’s low amylase activity, many of these beverages are often produced with the aid of amylase-producing lactobacillus (10) or similar microbes, which contributes a unique sour flavor (8). In -based brews that do not use these microbes, amylase enzymes must often be supplemented with the addition of other grain adjuncts or microbially derived enzymes to achieve liquefaction and saccharification (11,12). The addition of these elements can be cost prohibitive (13) and, when using supplemental grain adjuncts like barley, introduce gluten to the product. If sorghum accessions can be identified that display significant activity of beta or alpha amylase, they can be incorporated into breeding programs and used to improve commercial materials.

28 In efforts to identify superior malting sorghums, the presence of amylase inhibitors found in the grain must be considered due to their negative impact on starch hydrolysis. While there are many compounds endogenous to sorghum that have been shown or are speculated to inhibit the activity of these enzymes (14,15,16), tannins and phenols are well documented amylase inhibitors. These compounds are common among many sorghum accessions and can be produced in great quantities (17,18). While these compounds may have a variety of positive health effects in humans (19), phenolic extracts containing tannins have been shown to have inhibitory effects on porcine pancreatic amylases, human salivary amylases, and enzymatic activity of sorghum malts in aqueous solutions (14,20,21,22,23). This repressive activity negatively affects the function of amylases on starches and yeast metabolism (24), impacting fermentation success.

Breeding efforts may be able to reduce these compounds among accessions which display also high amylase activity, potentially further increasing their amylase activity.

Alternatively, sorghums could be developed to make malted foodstuffs that incorporate these health beneficial compounds while still maintaining elevated amylase activity necessary for their production.

Thus, an evaluation between amylase and phenols/tannins was made due to their impact on product quality.

A secondary consideration in identifying superior malting sorghums is grain color.

Several traditional beverages made with sorghum malts are made with preference to those

29 of specific color of the outer pericarp (25,26,27,28). For example, specific varieties of baijiu, a Chinese liquor, are made with red sorghums (26), as is dolo, an opaque African beer of low alcohol content (25). Furthermore, studies which have reported color in amylase evaluations have found several red and black sorghum accessions that display higher levels of alpha and beta amylase than their peers (3,28). Similarly, studies evaluating specific varieties of sorghum for their brewing quality, including amylolytic activity, show appreciable diastatic power (for sorghum) in red accessions (30,31).

Cultural preference among sorghums is not exclusive to red grained types, nor are they the only ones to display significant amylolytic activity. However, such observations prompted an investigation into the potential linkage between color and amylase activity.

Few, if any, studies have looked directly for an association between these two phenotypes, color and amylase activity, in a diverse collection of accessions, as done here. Such a connection could direct genetic investigations to loci controlling grain pigmentation.

Furthermore, many previous studies evaluating sorghum malt amylase activity have used either few samples, samples not grown under controlled conditions, and/or used material with limited diversity (2,32,33,34,35,36). For example, while a study by

Dicko et al. (2006) evaluated 50 accessions, the evaluated material was chosen for their purported culinary use, and not diversity. While a range of enzyme activities are presented in such studies, the inferences that can be made about material outside the study are constrained. To overcome these shortcomings, this study examined accessions from the grain sorghum diversity panel (37,38) grown in the same natural field

30 environment. Lines were selected to maintain representation of the original panel’s composition based upon criteria including materials’ geographic origin, genetic diversity, racial classification, visual phenotypic diversity, use as a genetic model, or importance in sorghum breeding. This design, with diverse accessions grown under identical conditions, allows for the identification of common features among lines that correlate with enzymatic activity.

These features can discriminate between the large amount of understudied sorghum accessions to target material that are likely to harbor similar amylolytic activity, which provides an avenue to direct efforts to identify superior malting sorghums.

Materials and methods

Plant materials and harvest

Accessions from the grain sorghum diversity panel (37,38) were grown in 2× replicated randomized complete block design of two row plot, 3m length, and a row spacing of 0.762m. in loamy sand soil. The experiment was planted on May 16, 2019 in

Florence, SC at the Clemson University Pee Dee Research and Education Center. Plant density was approximately 130,000 plants ha−1 and plots were irrigated as needed to create a favorable production environment for characterizing sorghum grain quality traits.

These accessions can be found in table B-3.

Primary panicles (i.e., panicles of the main culm) used in this study were randomly selected within each plot and harvested at physiological maturity. Anthesis date, the date at which half of the plants in plot had 50% anther emergence, was recorded

31 for each accession to ensure uniform collection at physiological maturity between accessions. To avoid the influence of border effect, plants at the beginnings and ends of rows were not harvested.

Near infrared spectroscopy evaluation of tannins and phenols

Tannins and phenolics from samples ground to 1-mm particle size were evaluated with near infrared spectroscopy performed in 2013, 2014, and 2017 as described elsewhere (18,39). Traits were measured with a DA 7250 NIR analyzer (Perten

Instruments). To calibrate the Perten DA 7250, 100 samples from the grain sorghum diversity panel were evaluated using wet chemistry performed at Dairyland Laboratories,

Inc. (Arcadia, WI) and the Quality Assurance Laboratory at Murphy-Brown, LLC

(Warsaw, NC). Both compound class concentration levels are displayed in their Catechin

Equivalent per gram (CE/g) and Gallic Acid Equivalent (GAE/g) for accession collected upon as displayed in table B-3.

Seed surface sterilization and treatment

In order to control against fungal or bacterial amylase contamination of the malts, each 5g sample of clean, unbroken grain was surface sterilized and treated with 70% ethanol for two minutes, followed by a Captan fungicide treatment for three hours in a rotary shaker at 250 rpm, and a 25% bleach treatment (8.25% sodium hypochlorite solution) with 3 drops of Ajax soap for 20 minutes. Sterilized seeds were rinsed five times with sterile DI water in a sterile laminar flow hood and transferred to magenta boxes.

32

Malting conditions

Two replicates of each accession were steeped at 23°C for 24 hours in a 0.1% formaldehyde solution at pH 6.5, drained, rested, and then steeped for an additional 24 hours in fresh formaldehyde solution as adapted from other studies (22–24). After steeping, seeds were spread uniformly across the malting container by gentle shaking and germinated for five days at 30°C. Ungerminated seeds were discarded and the malted grains were dried at 50°C for 24 hours in a 750F incubation oven (Fisher Scientific).

Malt amylase content evaluation

Malt roots and shoots were removed using a fine steel mesh strainer, and samples were ground using a GenoGrinder 2000 for 2 minutes at 1750 rpm in 15 mL jars with two steel ball bearings. Amylase content was determined using 0.25g of ground malt, as is, from each accession and the commercially available Megazymes

(https://www.megazyme.com/) Ceralpha and Betamyl assays with an extended 1-hour extraction and 30-minute incubation.

Spectrographic readings for each accession were determined using a Beckman

Coulter DU730 scanning at 400 nm and normalized to readings obtained from barley malt at known enzymatic content provided by the manufacturer.

Statistical analyses

JMP 14.3 pro (https://www.jmp.com/) was used to perform ANOVA analyses, t- tests (LSD), and F-tests for analysis of regions, race, and pericarp color compared to alpha and beta amylases. Additionally, JMP was used to generate graphic outputs, linear

33 models, and lack of fit analyses (F-tests) of phenolic and tannin content versus alpha and beta amylase. All levels of significance for methods used were at the P < 0.05 level.

Results and discussion

Amylase content in evaluated sorghums

A significant range of alpha amylase content was observed throughout the diverse accessions, ranging from 5.5 U/g to 248.2 U/g. The line SC 1057 (a short stature, photoperiod insensitive sweet-sorghum conversion of Grassl) showed alpha amylase content (mean 245.2 U/g) that was statistically similar to that of the barley control used in the study, and exceeded levels found in commercial malting barleys of other studies (40).

Conversely, accessions SC 1218 (a white pericarp, tannic sorghum) and No. 4 Hadoui (a grain nested association mapping panel parent) showed the statistically lowest observed mean alpha amylase content. The mean alpha amylase levels observed in this study’s sample was 64.8 U/g, below the mean of other studies evaluating wild (94.3 U/g) (41) and commercial barleys (165.5 U/g) (40) assayed using the same CERALPHA assay.

However, a greater range and higher mean of alpha amylase activity was observed than those reported in other studies of sorghum (2,42) such as by Beta et al. (1995). which reported a range of 25 to 183 U/g. This is likely due to the greater diversity in this study's sorghum material and, potentially, differences in malting method as compared to these previous experiments, though subsequent studies would have to evaluate directly to confirm. While other studies have reported other sorghums with alpha amylase content that contend with commercial barleys (34), the levels observed here are substantially

34 above those found in studies using analogous assay methods (2,42). This observation is noteworthy due to the implications for malting and other sorghum products. Not only does high activity of alpha amylase rapidly release dextrins during mashing that beta amylase can utilize, but its liquefaction activity is essential for foodstuffs with high soluble content such as syrups and weaning products. Utilizing accessions which present high alpha amylase, such as SC 1057, to develop improved lines could thus improve food security in developing countries (43). Furthermore, crossing high alpha amylase accessions with high beta amylase accessions identified in other studies (44,45), may help breeders produce improved cultivars or hybrids with a significantly reduced need for amylolytic enzyme supplementation.

Beta amylase activity observed in this study was lower than that observed by other similar studies (2,31), with observations ranging from 2.8 U/g to 31.7 U/g. These studies demonstrated ranges from 16 U/g to 57.2 U/g (31) or 11 to 41 U/g (2). The lower overall mean levels of beta amylase activity may have been the result of the malting method used in this study. Variations in the malting process have been linked to differences in amylase activity in other sorghum studies (3,31,32). However, beta amylase levels showed no statistical difference between replicated , validating the observed results as true for this method. The 30°C germination temperature used in this study may have favored the development of alpha amylase over beta amylase as observed in other studies (46). Alternate malting methods may reveal elevated beta amylase levels for the accessions used, such as those used by Letsididi et al. (2008).

Among the material evaluated for beta amylase, SC 1057 again showed the greatest

35 enzymatic activity, with a mean of 31.7 U/g. Conversely, HC 6028 (a white pericarp and non-tannic accession) showed the lowest beta amylase activity at 2.8 U/g. For comparison, the barley control malt displayed beta amylase activity of 19.2 U/g, however commercial malts typically display beta amylase values > 50 U/g (2). While malts of SC

1057 and SC 720 (21.1 U/g) display beta amylase activity greater than the control, they are below what is commercially desired for self-sufficient saccharification. Fig. A-1 highlights the mean values for both beta and alpha amylase of the accessions evaluated, and a connecting letters report for significant difference between their means can be found in table B-3.

In examining the commonly evaluated sorghum genetic resources, with the exception of the red-pericarp accession Shan Qui Red (PI 656025), the grain nested association mapping parents (SAP-173, No. 4 Hadoui, P898012, RTx430, Macia,

Ajabsido, No. 69 Mashela Tinguish Warakul Ethiopi, Msumbji SB 117) and genomic reference line BTx623 had significantly lower alpha amylase activity (p value 0.0089) than the remainder of the diverse samples evaluated.

Even Shan Qui Red, with a mean alpha amylase content of 126.11 U/g, is still well below the observed 245.20 U/g in SC 1057. Additionally, while observed overall beta amylase levels were low, these accessions also displayed a limited range of 7.6 U/g to 12.8 U/g, and displayed low variation compared to the greater collection (two tailed F- test p value 0.0002). These observations undermine the usefulness of these resources in genetic mapping of amylase activity. The lack of material exhibiting sufficient amylase activity among these commonly used genetic resources may hamper future efforts to

36 identify key genetic regulators of the trait. However, the converted line of Grassl, SC

1057, shows significant activity of alpha and beta amylase compared to the sample evaluated. Serendipitously, Grassl is a recurrent female parent in the sorghum bioenergy nested association mapping panel (unpublished, described in 42), future efforts to map loci associated with amylolytic activity may be best served using these genetic resources.

Evaluations on the remaining founders of this population (unpublished), using the methods described here, show appreciable variation for amylase activity supporting the use of this population for genetic mapping.

Geographic and racial distribution of amylase content

As sorghum has migrated with humans, it has fulfilled diverse needs and has been subjected to different selective pressures by both their habitats and humans, including for malting (48,49). As such, an evaluation of sorghums from diverse regions revealed variations between regions that can guide subsequent evaluations of additional sorghums.

To evaluate the geographic distribution of amylase content, accessions were categorized into “Center of Diversity” (being the region considered to be the center of sorghum diversity), “East Africa”, “Southern Africa”, “West/”, “South ”, “United

States”, and “International” for accessions originating in other regions. The distribution of materials is highlighted in Fig. A-5, and the classifications of regionality of Africa is noted in Fig. A-2. Accessions originating from and East Africa had statistically greater alpha amylase content than those in Western/North Africa or

Southern Africa, while sorghums from the Center of Diversity encompassed the greatest range of alpha amylase content (8.09 to 248.29 U/g) and were not statistically distinct

37 from those of other regions. Given the expected genetic diversity of material from the region and in material sampled, the wide range of alpha amylase activity observed in this study appears to be representative of the broader range of sorghum activity. For beta amylase, a statistically lower mean activity level among accessions from Southern Africa

(p value 0.037) and West/North Africa (p value 0.007) was observed when compared to the other accessions. No other associations were found, which is likely due to the overall low levels observed in this study and in sorghum at large. These results indicate that in addition to germplasm from the center of diversity, sorghums from East Africa and South

Asia are potential sources that may exhibit elevated alpha amylase content in their malts.

Subsequent studies should incorporate additional and a wider range of materials, allowing for finer geographic mapping of the phenotypes.

———

As phenotypic diversity within grain traits correlates with sorghum racial classifications

(48,50,51), an evaluation was performed to see if amylase content of malts did as well.

Racially, bicolors showed a statistically significantly greater in both alpha and beta amylase activity in their malts (p value 0.0026 & 0.0002, respectively) compared to other racial groups, while kafirs showed significantly (p value 0.0069) lower beta amylase activity in the studied sorghums.

When only comparing accessions originating from Africa, accessions showed significantly lower levels (p value 0.0221) of alpha amylase than other African accessions, while bicolors still had higher levels (p value 0.0114).

38 In exploring potential explanations of these trends in both regionality and race, the migration of people and origins of the races of sorghum should be taken into consideration. The more “primitive” and diverse bicolors, from which other races have been postulated to derive (48,49), may have experienced less evolutionary pressure for grain storage conditions, among which conservative regulation of amylases would be preferred to prevent grain mold. This also may explain the levels observed in guineas, as the high-rainfall regions they are adapted too (48) are environments which encourage the development of grain mold. potential explanation for these results is that Bantu tribes historically perpetuated the tradition of making beer and malted foodstuff with sorghum as they travelled south, which may also explain the increased alpha amylase activity among accessions from East Africa. These migration patterns may also explain why guinea accessions in Western Africa are observed to have lower levels of both amylases, as their presence in West Africa predates the Bantu migration southward (48,52).

Ultimately, subsequent evaluations will have to be conducted among global sorghum collections to verify the extent to which origin and race impact amylase characteristics.

Nevertheless, these data may allow subsequent studies to discriminate between the large amount of understudied sorghums when attempting to identify those with superior amylase characteristics.

Pericarp color and amylase content

Several studies have used or found specific red (or black) sorghums which produce high levels of alpha and/or beta amylase (2,31). Thus, an examination of seed color and amylase activity was conducted to identify possible associations across this

39 diverse population. Red- pericarp sorghums showed significantly greater alpha amylase and beta amylase activity than other pericarp colors (Fig. A-3). Given that white and yellow are favored grain colors in the United States for food (53), much of the material from the United States is lacking in enzyme activity in their malts. These findings are also compelling given that red-pericarp accessions were found to be more vigorous in germination among near isogenic lines developed to examine the agronomic influence of color (53). Among these loci that would be variable in the near isogenic lines, the epistatic R and Y loci, both responsible for the red color of the pericarp. The lack of elevated amylase activity in the yellow pericarp accessions studied indicates that the Y locus is likely not associated with elevated amylase levels (54). As amylase activity is essential in germination, the observations presented here and, in the study, examining these near isogenic lines suggest that pericarp color and genetic regulators for amylase content are linked. Due to the apparent association between red-pericarp pigmentation and amylase content in sorghum malt, the surrounding genetic region for the R locus that controls red color should be examined in future studies.

Tannins, phenolics, and malt amylase levels

Previous studies have directly evaluated the content of phenolic extracts with wet chemistry approaches (23,30,55), while the near infrared spectroscopy methods, used here, relies on indirect analysis of past samples with known analyte content. Thus, calibration curves obtained for these known activities can be applied to infer the content in unknown samples. This method has been applied to evaluate sorghum content of tannins and phenolics with sufficient acuity for genetic mapping (18). Using near infrared

40 spectroscopy, a significant lack of fit (F- test) to expected linear models was found between either phenolics or tannins to either amylase activity of the malted accessions

(Fig. A-4). With the expected negative impact that these compounds have on amylase activity seen in aforementioned studies, the very weak and slightly positive linear association between these compounds and enzymes observed in this study is notable.

This lack of an expected negative association between phenols/tannins and amylase activity indicates that the traits may largely segregate independently and/or be confounded by other factors which complicated the relationship. A significant association between red colored sorghums and tannins and phenols was observed and may be in part on such factor, however, these phenol and tannin levels observed here were analogous for accessions shared between other studies, such as Macia and Ajabsido (21,43).

Additionally, despite this association several red sorghums studied showed no detectable phenols nor tannins and still displayed a range of amylase activities, including SC 720,

SAP-225, and SAP-280. Interestingly, SC 1057 displayed the greatest levels of phenols

(14.6 GAE/g) as well as amylase activity among the sample.

However, several accessions displayed no detectable phenols using near infrared spectroscopy, yet displayed notable alpha amylase activity, such as Tx432 (139.8 U/g). In analyzing tannins, Red Amber showed the greatest phenol content (24.1 GAE/g) while containing appreciable amounts of both alpha (125 U/g) and beta (18 U/g) amylase compared to other accessions. Many other accessions did not show quantifiable tannin content. It may be possible that removal of these compounds through breeding from high amylase accessions will result in even greater amylase activity in sorghum malt-mashes.

41 For example, crossing high alpha amylase accessions, such as SC 1057, with those that exhibit elevated beta amylase and low content of inhibitory compounds may lead to a more optimal malt sorghum. Measured phenols and tannins for the accessions can be found in table B-2.

Conclusions

In this study, alpha and beta amylase activity of malt was examined in a diverse sampling of sorghum accessions obtained from the grain sorghum diversity panel grown in identical conditions. SC 1057 was identified as an accession with significant levels of alpha amylase in its malt, comparable with a commercial barley accession, and contained appreciable beta amylase among the accessions evaluated. As SC 1057 is a converted line of Grassl, it may be possible to use the sorghum bioenergy nested association mapping population to identify loci associated with amylase activity. This is fortuitous as much of the commonly used genetic resources in the sorghum community, including the grain nested association mapping population founders and genomic reference BTx623, show poor amylase activity. Additionally, a significant association between red-pericarp sorghums’ malts and amylase content was found within this sample.

Among races of sorghum, bicolors showed significantly greater levels of both amylases, guinea accessions showed lower amylase levels, and kafirs displayed low beta amylase activity.

Accessions examined in this study that originated from South Asia and East

African accession showed higher activity in alpha amylase while those of Southern and

42 West/North Africa showed the lowest of both enzymes. Subsequent studies will be necessary to further validate these findings. With this study’s malting method, no association was identified between amylases and phenolics nor tannins. This indicates that these two traits largely segregate independently, and thus may be bred apart in lines such as Red Amber, which displays elevated levels of each.

These findings demonstrate that the introduction of material with greater amylase activity would benefiting efforts to develop superior malting accessions of sorghum.

References

1. Bamforth, C. W. 2009. Current perspectives on the role of enzymes in brewing. J.

Cereal Sci. https://10.1016/j.jcs.2009.03.001

2. Beta, T., Rooney, L. W., and Waniska, R. D. 1995. Malting characteristics of

sorghum cultivars. 72 (6) :533–538

3. Dewar, J., Taylor, J. R. N., and Berjak, P. 1997. Determination of improved steeping

conditions for sorghum malting. J. Cereal Sci. https://10/fb48df

4. Agu, R. C., and Palmer, G. H. 1998. A reassessment of sorghum for lager-beer

brewing. Bioresour. Technol. https://10/brvg5b

5. Arendt, E., Dal Bello, F., and Bello, F. D. 2011. Gluten-free cereal products and

beverages. Elsevier.

6. Taylor, J. R. N. 1992. Mashing with malted grain sorghum. J. Am. Soc. Brew. Chem.

https://10/ccz8cj

7. Owuama, C. I. 1999. Brewing beer with sorghum. J. Inst. Brew. https://10/ggwqr7

43 8. Lyumugabe, F., Gros, J., Nzungize, J., Bajyana, E., and Thonart, P. 2012.

Characteristics of African traditional beers brewed with sorghum malt: A review.

16 (4)

9. Zheng, X.-W., and Han, B.-Z. 2016. Baijiu (白酒), Chinese liquor: History,

classification and manufacture. J. Ethn. Foods https://10/ggwqsg

10. Sawadogo-Lingani, H., Diawara, B., Traoré, A. S., and Jakobsen, M. 2008.

Technological properties of lactobacillus fermentum involved in the processing of

dolo and pito, West African sorghum beers, for the selection of starter cultures. J.

Appl. Microbiol. https://10/dv83hf

11. Taylor, J. R. N., Dlamini, B. C., and Kruger, J. 2013. 125th anniversary review: the

science of the tropical cereals sorghum, maize and rice in relation to lager beer

brewing. J. Inst. Brew. https://10/gfxjx3

12. Espinosa-Ramírez, J., Pérez-Carrillo, E., and Serna-Saldívar, S. O. 2013. Production

of lager beers from different types of sorghum malts and adjuncts supplemented

with β-amylase or amyloglucosidase. J. Am. Soc. Brew. Chem. https://10/gbd7tz

13. Palmer, G. H., Etokakpan, O. U., and Igyor, M. A. 1989. Sorghum as brewing

material. MIRCEN J. Appl. Microbiol. Biotechnol. https://10/b6j9ff

14. Bloch, C., and Richardson, M. 1992. The amino acid sequences of two 13-kda

alpha-amylase inhibitors from the seeds of Sorghum bicolor (l.) Moench. 5 (1):

27–30

15. Mulimani, V. H., and Supriya, D. 1993. Alpha amylase inhibitors in sorghum

(Sorghum bicolor). Plant Foods Hum. Nutr. https://10/bktcn8

44 16. Chandrasekher, G., Raju, D. S., and Pattabiraman, T. N. 1981. Natural plant enzyme

inhibitors. α-amylase inhibitors in . J. Sci. Food Agric. https://10/b9tg9s

17. Habyarimana, E., Dall’Agata, M., De Franceschi, P., and Baloch, F. S. 2019.

Genome-wide association mapping of total antioxidant capacity, phenols, tannins,

and flavonoids in a panel of Sorghum bicolor and S. bicolor × S. halepense

populations using multi-locus models. PloS One https://10/ggwqsq

18. Rhodes, D. H., Hoffmann Jr, L., Rooney, W. L., Ramu, P., Morris, G. P., and

Kresovich, S. 2014. Genome-wide association study of grain polyphenol

concentrations in global sorghum [Sorghum bicolor (l.) Moench] germplasm. J.

Agric. Food Chem. https://10/f6qkgq

19. Dykes, L., and Rooney, L. W. 2006. Sorghum and millet phenols and antioxidants.

J. Cereal Sci. https://10/d4rwq6

20. Kim, J.-S., Hyun, T. K., and Kim, M.-J. 2011. The inhibitory effects of ethanol

extracts from sorghum, foxtail millet and proso millet on α-glucosidase and α-

amylase activities. Food Chem. https://10/dsb67r

21. Mkandawire, N. L., Kaufman, R. C., Bean, S. R., Weller, C. L., Jackson, D. S., and

Rose, J. 2013. Effects of sorghum (Sorghum bicolor (l.) Moench) tannins on α-

amylase activity and in vitro digestibility of starch in raw and processed . J.

Agric. Food Chem. https://10/f4v359

22. Beta, T., Rooney, L. W., Marovatsanga, L. T., and Taylor, J. R. N. 2000. Effect of

chemical treatments on polyphenols and malt quality in sorghum. J. Cereal Sci.

https://10/fq55wg

45 23. Daiber, K. H. 1975. Enzyme inhibition by polyphenols of sorghum grain and malt.

J. Sci. Food Agric. https://10/fqmsdh

24. Mullins, J. T., and NeSmith, C. 1988. Nitrogen levels and yeast viability during

of grain sorghum containing condensed tannins. Biomass

https://10/bt8qb7

25. Abdoul-latif, F. M., Bassolé, I. H., Bassolé, I. H., Dicko, M. H., and Dicko, M. H.

2013. Proximate composition of traditional local sorghum beer “dolo” manufactured

in Ouagadougou. Afr. J. Biotechnol. https://10.4314/ajb.v12i13.

26. Tan, G., Hu, Y., Huang, Y., Liu, H., Dong, W., Li, J., Liu, J., Peng, N., Liang, Y.,

and Zhao, S. 2020. Analysis of bacterial communities in pit mud from Zhijiang

Baijiu distillery using denaturing gradient gel electrophoresis and high-throughput

sequencing. J. Inst. Brew. https://10/ggwqsf

27. Kayodé, A. P. P., Hounhouigan, D. J., Nout, M. J. R., and Niehof, A. 2007.

Household production of sorghum beer in Benin: Technological and socio-economic

aspects. Int. J. Consum. Stud. https://10/fcg295

28. Kayode, A. P. P., Adegbidi, A., Hounhouigan, J. D., Linnemann, A. R., and Nout,

M. J. R. 2005. Quality of farmers’ varieties of sorghum and derived foods as

perceived by consumers in Benin. Ecol. Food Nutr. https://10/d8dxqk

29. Beta, T., Corke, H., and Taylor, J. R. N. 2000. Starch properties of Barnard red, a

South African red sorghum variety of significance in traditional African brewing.

Starch - Stärke https://10/dt853b

30. Tawaba, J.-C. B., Tshiala, H., Kibal, I., Buetusiwa, T., and Minengu, J. de D. 2015.

46 Effects of phenolic compounds on the hydrolysis of red sorghum starch by extracted

red sorghum malt α- and β-amylases. Starch - Stärke https://10/f276zt

31. Taylor, J. R. N., and Robbins, D. J. 1993. Factors influencing beta amylase activity

in sorghum malt. J. Inst. Brew. https://10/ggwqsm

32. Adefila, O. A., Bakare, M. K., and Adewale, I. O. 2012. Characterization of an α-

amylase from sorghum (Sorghum bicolor) obtained under optimized conditions. J.

Inst. Brew. https://10.1002/jib.11

33. Aisen, A. O., and Muts, G. C. J. 1987. Micro-scale malting and brewing studies of

some sorghum varieties. J. Inst. Brew. https://10/ggwqst

34. Dufour, J. P., Melotte, L., Mélotte, L., Srebrnik, S., Melotte, L., and Srebrnik, S.

1992. Sorghum malts for the production of a lager beer. J. Am. Soc. Brew. Chem.

https://10/dz5ht2

35. Ogu, E. O., Odibo, F. J. C., Agu, R. C., and Palmer, G. H. 2006. Quality assessment

of different sorghum varieties for their brewing potential. J. Inst. Brew.

https://10/ggwqs2

36. Ratnavathi, C. V., and Ravi, S. B. 1991. Effect of different durations of steeping and

malting on the production of alpha-amylase in sorghum. J. Cereal Sci.

https://10/d5t9hd

37. Boyles, R. E., Pfeiffer, B. K., Cooper, E. A., Rauh, B. L., Zielinski, K. J., Myers, M.

T., Brenton, Z., Rooney, W. L., and Kresovich, S. 2017. Genetic dissection of

sorghum grain quality traits using diverse and segregating populations. Theor. Appl.

Genet. https://10/f94sq2

47 38. Casa, A. M., Pressoir, G., Brown, P. J., Mitchell, S. E., Rooney, W. L., Tuinstra, M.

R., Franks, C. D., and Kresovich, S. 2008. Community resources and strategies for

association mapping in sorghum. Crop Sci. https://10/b8h7sj

39. Dykes, L., Hoffmann, L., Portillo-Rodriguez, O., Rooney, W. L., and Rooney, L. W.

2014. Prediction of total phenols, condensed tannins, and 3-deoxyanthocyanidins in

sorghum grain using near-infrared (NIR) spectroscopy. J. Cereal Sci.

https://10/f59gn3

40. Schnitzenbaumer, B., Kaspar, J., Titze, J., and Arendt, E. K. 2014. Implementation

of commercial and sorghum flours in brewing. Eur. Food Res. Technol.

https://10/f5tjqb

41. Cu, S. T., March, T. J., Degner, S., and Eglinton, J. K. 2016. Identification of novel

alleles from wild barley for the improvement of alpha-amylase and related malt

quality traits. Plant Breed. https://10/f9fdxk

42. Letsididi, R., Bulawayo, B., Kebakile, M., and Ezeogu, L. I. 2008. Evaluation of

indigenous Botswana sorghum cultivars with respect to their diastatic power, α-

amylase, β-amylase, and limit dextrinase potentials for malting. J. Am. Soc. Brew.

Chem. https://10/bg9pqr

43. Dicko, M. H., Gruppen, H., Traoré, A. S., Voragen, A. G. J., and van Berkel, W. J.

H. 2006. Sorghum grain as human food in Africa: Relevance of content of starch and

amylase activities. Afr. J. Biotechnol. https://10.4314/ajb.v5i5.

44. Ogbonna, A. C. 1992. Developments in the malting and brewing trials with

sorghum. World J. Microbiol. Biotechnol. https://10/dg958r

48 45. Okungbowa, J., Obeta, J. a. N., and Ezeogu, L. I. 2002. Sorghum β-amylase

production: Relationship with grain , steep regime, steep liquor composition

and kilning temperature. J. Inst. Brew. https://10/ggwqs4

46. Bwanganga, J.-C. T. 2015. Effect of cold shock on the enhancement of β-amylase

activity during malting and malt processability for a red sorghum intended for

brewing use. J. Inst. Brew. https://10/ggwrvq

47. Boyles, R. E., Brenton, Z. W., and Kresovich, S. 2019. Genetic and genomic

resources of sorghum to connect genotype with phenotype in contrasting

environments. Plant J. https://10/ggwqsn

48. Westengen, O. T., Okongo, M. A., Onek, L., Berg, T., Upadhyaya, H., Birkeland, S.,

Khalsa, S. D. K., Ring, K. H., Stenseth, N. C., and Brysting, A. K. 2014.

Ethnolinguistic structuring of sorghum genetic diversity in Africa and the role of

local seed systems. Proc. Natl. Acad. Sci. https://10/f6jf4n

49. Morris, G. P., Ramu, P., Deshpande, S. P., Hash, C. T., Shah, T., Upadhyaya, H. D.,

Riera- Lizarazu, O., Brown, P. J., Acharya, C. B., Mitchell, S. E., Harriman, J.,

Glaubitz, J. C., Buckler, E. S., and Kresovich, S. 2013. Population genomic and

genome-wide association studies of agroclimatic traits in sorghum. Proc. Natl. Acad.

Sci. https://10/f4m576

50. Brown, P. J., Myles, S., and Kresovich, S. 2011. Genetic support for phenotype-

based racial classification in sorghum. Crop Sci. https://10/cmkptg

51. Harlan, J. R., and Wet, J. M. J. de. 1972. A simplified classification of cultivated

sorghum1. Crop Sci. https://10/drbn68

49 52. Smith, C. W., and Frederiksen, R. A. 2000. Sorghum: Origin, history, technology,

and production. John Wiley & Sons.

53. Pedersen, J. F., and Toy, J. J. 2001. Germination, emergence, and yield of 20 plant–

color, seed–color near-isogenic lines of grain sorghum. Crop Sci. https://10/ccbnb4

54. Clará Valencia, R., and Rooney, W. 2009. Genetic control of sorghum grain color.

INTSORMIL Presentations. 10. https://digitalcommons.unl.edu/intsormilpresent/10

55. Bvochora, J. M., Danner, H., Miyafuji, H., Braun, R., and Zvauya, R. 2005.

Variation of sorghum phenolic compounds during the preparation of opaque beer.

Process Biochem. https://10/ft24s2

50 CHAPTER THREE

FUTURE DIRECTIONS

An alternative to improve beta amylase in sorghum malts

Low beta amylase content is observed for sorghum malts in this study as well as others (Taylor et al., 1993; Beta et al., 1995). Compared to these studies however, our results were even lower than expected, which may be due to the malting method.

Interestingly, alterations to the malting method may be the best way to improve sorghum amylase productivity. There may ultimately be no manner of increasing beta amylase to the levels that are observed in barley due biological differences between development in each crop. Unlike barley, which produces high levels of both amylases as it germinates, sorghum does not. This may be due production of these enzymes “as needed” in sorghum rather than, effectively, all at once as in barley (Pagano et al., 1997). Physiological differences, such as where these enzymes are produced in the seed, have also been implicated to be causal, in part, for these differences (Aisien & Palmer, 1983). In barley, amylases are produced in the aleurone inside of the seed coat, while in sorghum they are produced in the scutellum (Aisien & Palmer, 1983).

Any improvement in beta amylase content (or alpha amylase for that matter) for the production of beverages ultimately can lead to reduced prices for commercial producers. With greater endogenous amylase content deriving from sorghum in the wort, less adjust barley or supplemental enzymes must be added (Taylor et al., 2013). One

51 method which has been explored is alterations in the malting method. Lower germination temperature and shorter germination times have a positive impact on beta amylase activity but appear to reduce alpha amylase activity, though it appears to be dependent on accessions used (Agu & Palmer, 1997). A full factorial design that compares these conditions among a diverse collection of sorghums may reveal material which displays appreciable activity for both amylases under specific conditions. This idea, while not novel (Taylor et al. 1993), has yet to be executed among more than a handful of accessions. This may ultimately be the best way to reduce costs and improve malting quality, through optimizing a method, and developing sorghums which perform well within them.

Potential, evaluation, and incorporation of malting germplasm

The current sorghum collections used for genetic mapping or by commercial breeders in the United States were not constructed for improvement of malting, in part, due to sorghum's low historic use for the purpose in the country. Even ICSV 400, a sorghum accession released for malting purposes (Murty et al. 1998), showed low amylase levels in our study. Thus, to improve both genetic mapping, marker development, and breeding efforts for malting sorghums, additional germplasm needs to be screened and incorporated into existing programs. As discussed in chapter two there are several potential sources of diversity both geographically and racially among global sorghum germplasm that can be used for this goal. However, based upon the literature,

52 there are other regions that should be investigated for germplasm in future efforts more thoroughly.

Due to the long history of making baijiu in China, future studies should be designed to investigate a greater variety of sorghums from the region. These investigations are likely to reveal higher quality malting sorghums based on the continued selection for beverage making among Chinese germplasm (Yan et al. 2018). One limitation of our study was that our grain sorghum diversity panel has very few accessions originating from China (Boyles et al., 2017). However, obtaining and studying international materials is usually challenging considering governmental restrictions and transfer agreements (or lack thereof) between countries. Many countries, China included, view their germplasm as a resource to be used for the country's benefit and guard the value by limiting sharing with other countries (Frison et al., 2011). With current trade wars between China and the US, such restrictions are unlikely to be lifted. It may be more productive to obtain germplasm from international partners in South Asia, such as

ICRISAT, despite the high malting potential of Chinese sorghum accessions.

Sorghum material may also be difficult to obtain because much of the historic diversity in these regions has potentially been lost. Local growers often abandon landraces and traditional varieties in favor of high yielding commercial cultivars with desirable agronomic traits (Lightbourne, 2016; Plucknett et al. 1987). Since World War

II, sorghum landraces have been abandoned in favor of high yielding varieties in regions where those landraces were once indigenous (Plucknett et al. 1987). With landraces being

53 replaced and international cooperation impacted, the available time frame to reobtain traditional sorghums with superior malting traits may be rapidly closing.

Obtaining sorghum accessions with superior malting characteristics will aid in breeding and genetic mapping efforts. Mapping of loci associated with malting quality is important for both developing molecular markers to aid breeders and for identifying causal genetic variation. There are several genomics tools and techniques by which this can be done, each with advantages and disadvantages.

Linkage mapping with recombinant inbred lines

The first mapping design discussed here uses recombinant inbred lines (RILs), which serve as a linkage mapping population to discover quantitative trait loci (QTL).

QTLs are loci that correlate with quantitative variation in a phenotype within a population. QTLs are identified through mapping using molecular markers that are segregating between individuals in a population and are associated with the phenotype.

Markers that are found in association with QTLs can be used by breeders to make selections, by genotyping at the QTL locus, and selecting individuals with the associated markers. These QTLs can be targeted in subsequent studies to determine the causal genetic variations for the phenotypic variation, for example for amylases (Myles et al.

2009; Pollard, 2012).

RIL populations are produced by crossing two or more phenotypically divergent parents and then selfing the offspring until they are homozygous for almost all loci.

These lines have alternating parental chromosomal contributions, or haplotypes, of

54 varying lengths which are determined, typically, by genotyping molecular markers interspersed throughout the genome.

These chromosomal contributions dictate the variation in phenotype that is observed between these lines. Thus, the parental genetic contribution to the variance of the phenotype between lines at such loci can be estimated, and thus QTLs for the trait can be identified (Pollard, 2012).

RILs have previously been employed in sorghum to map for grain composition

(Boyles et al., 2017a), drought tolerance (Phuong et al., 2019), and grain yield components (Boyles et al., 2017b). They can likely be employed to map QTLs for amylase productivity because amylases are quantitatively inheritance, thus, likely to show transgressive segregation. A potential RIL population to map amylase could be created using BTx623, a genetically well characterized and low amylase producing line, and SC

1057, the high amylase line we found in our study.

Mapping with RILs provides several advantages over using association mapping techniques. Because true random mating does not exist in nature, there are complex population structures and interrelatedness that exist in any such population and lines derived therefrom (Myels et al. 2009). Thus, when using traditional diversity panels and association mapping based techniques, which rely on markers across the genome and historic recombination, many such markers may falsely appear to be relevant when genotypes covary with phenotypes. These challenges may be overcome by using RIL populations, which use recent recombination events to provide genetic variation for mapping. Furthermore, RILs may often have a smaller phenotyping burden than other

55 mapping populations due to having fewer numbers of constituent individuals needed to sufficiently map loci associated with a trait (Myels et al. 2009). As a result, for time consuming, expensive, or difficult to phenotype traits, the use of RILs can be an effective mapping method.

However, there are also drawbacks to linkage mapping with RILs. There are significant time and cost requirements to develop these populations, ensuring the offspring lines are sufficiently homozygous for linkage mapping. Furthermore, because the recent recombination events can result in large regions of inheritance, they often contain numerous loci (Myles et al. 2009). For this reason, RILs might be a less than optimal method of identifying the causal genetic basis of the observed variation in either amylase activity but can be sufficient to identify markers for QTLs to inform breeding decisions. Furthermore, as discussed below, if established populations exist that have constituents that vary for the trait, the development of a RIL population is redundant.

Genetic mapping with nested association mapping populations

Another method for the genetic mapping of sorghum malting traits, amylase included, can be the use of nested association mapping (NAM) populations. These populations are designed to leverage both recent and historic recombination events to break up established haploblocks and improve genetic mapping resolution over mapping with diversity panels or RILs alone (Bouchet et al., 2017). This largely eliminates the disadvantages of linkage and association mapping, while retaining many of the benefits of both. In comparing these populations to RILs, NAMs are often larger and thus have a

56 significantly greater phenotyping burden. Additionally, the growing seasons required to develop a NAM population are more or less equivalent to that of a RIL population, and thus are theoretically no faster to develop.

Construction of NAM populations involves the selection of several genetically and phenotypically diverse male parents and crossing them to one recurrent female parent

(McMullen et al., 2009). The genetic diversity between the male parents provide historic genetic recombination, while the controlled crosses to the recurrent female break up established regions of high linkage disequilibrium. Following the initial cross, the offspring lines of this generation are selfed for several generations to attain homozygosity via single seed descent, similar to developing a RIL population.

Leveraging existing NAM populations for the mapping of amylase content and other malting characteristics can facilitate rapid discovery of molecular markers associated with variation in these traits. In sorghum, the NAM populations have been constructed to map grain traits (Bouchet et. al. 2017) as well as bioenergy traits

(Unpublished). Multiple parents from the grain NAM including Shan Qui Red, SAP-173,

No. 4 Hadoui, P898012, RTx430, Macia, Ajabsido, No. 69 Mashela Tinguish, Warakul

Ethiopi, and Msumbji SB 117, were included in our study to evaluate the phenotypic variation in this commonly used genetic mapping resource. To our surprise, the greatest observed level of amylase was seen not in a grain NAM parent, but in SC 1057, which is the converted accession of the bioenergy NAM parent Grassl.

As Grassl is already established as a parent in the bioenergy NAM (unpublished), this population could be used, saving the time needed to develop a novel RIL population

57 for genetic mapping. Significant variation in alpha amylase among the NAM founders was discovered as part of this work (table B-3), as such it is quite likely that this population can be used to map the phenotype. However, little variation for beta amylase was observed. Phenotyping the NAM larger populations to map for beta amylase associated loci may not be fruitful and may require some other population that may provide the mappable variation.

As the bioenergy NAM appears to be a promising population of study for amylase activity, there may be additional connections between important bioenergy traits and grain characteristics associated with malting. More specifically, directed carbon use, transport, and partitioning in the plant as a whole is one defining difference between bioenergy sorghums and other sorghums. For example, the variants in vacuolar ion transporters of Rio, Wray, Leoti, and Sugar Drip sorghum cultivars are associated with accumulation of non-structural carbohydrates (Brenton et al., 2016). Two phenotypes associated with efficient carbohydrate use and that have value in bioenergy production are also observed in grain of converted Grassl: early season vigor (of red sorghums) and greater productivity of alpha amylase. In traditional grain sorghums, genes controlling malting characteristics might have been selected against as they contribute to poor grain storage, flavor, or yield characteristics. Yet, grain sorghums have been used as genetic foundations for evaluating sorghum malting characteristics in most studies (Beta et al.

1995).

The bioenergy sorghums, as a whole, have not had such selection and can act as a source of variation for these traits. The proposed study screening the bioenergy NAM

58 parents could provide further evidence of such a connection and potentially establish that bioenergy sorghums might harbor quantitative variation that can be capitalized for malting improvement.

Causal gene identification

If mapping efforts identify markers or loci associated with elevated amylase content, there are multiple advantages to then identifying the causal variation. First, there is a gap in the understanding of genetic variation associated with amylase as compared to other crops.

Comparisons in the genetic variations of the enzyme between barley, rice, and wheat have revealed differences in each species’ gene copy numbers, sequence variations, and conserved promoter sequence motifs each putatively contribute to saccharification ability (Zhang & Li, 2017). Understanding genetic variations of sorghum that are responsible for superior amylase production can further the understanding of what factors, across species, regulate the trait.

Additionally, identifying responsible gene variants allows for the transgenic introduction of productive sorghum amylases to species that do not exhibit the trait nor exhibit natural variation. Alpha amylase content in maize biomass is to such a low level that industrial bioenergy efforts to transgenically introduce a highly active thermostable alpha amylase have been undertaken (Eichler, 2001). As has been previously proposed

(Kumar et al., 2005), leveraging the comparatively thermostable alpha amylase of sorghum (Adefila et al., 2012; Kumar et al., 2005), with an understanding why specific

59 sorghums are more productive, provides an alternative means to supply the enzyme for industrial applications. Finally, knowing the causal gene sequence would allow for the transgenic transfer of sorghum amylases to other crops and avoid the use of microbial alpha amylases, which are more expensive to deregulate.

The existing literature points to one possible a-priori genomic region that should be investigated. In our study we observed an association between amylase productivity and red seed color. One locus, the R locus, is a major regulator of red pericarp color in sorghum. When evaluating putative genes around the R locus, however, there are no predicted amylase genes.

However, a study comparing nearly isogenic lines of varying seed coloration observed more vigorous germination in red genotypes as compared to other colors.

Subsequent studies evaluating if this effect was due to resistance to pathogenic microbes, as had been postulated, found no difference in their ability to ward off pathogens (Funnell

& Pendersen et al., 2006). Instead, it is probable that genetic variation in a regulator of amylase is in association with the R locus and is responsible for increased amylase production in the seed. This would explain the more vigorous germination observed in the study, as rapid mobilization of carbohydrates would allow for rapid germination.

However, this would have to be verified through a candidate gene validation methods, of which several potential methods are discussed below.

Identification of the causative gene variation, whether targeted by a-priori or genetic mapping approaches, can be done in a surrogate organism. Using sequence homology to known proteins or regulatory sequences of other organisms associated with

60 amylase, loci for gene validation can be selected from the initial region of interest. These target loci from both a highly amylase (alpha or beta) productive accession and a non- productive accession should be amplified by PCR. These PCR amplicons should then be sequenced allowing for the fine sequence comparison between the two genes. If variations do exist between these sequences, then this locus would move forward in the pipeline. These putative amylase associated genes would be incorporated into transformation constructs that can be transferred to a model, such as yeast, for further study. Then a comparison of the amylase productivity between yeast transformed with the variant gene and a control transformed with a non-variant gene can be evaluated. This experiment would support the role of this variant gene impacting amylase productivity if elevated levels in yeast expressing the highly productive variant are observed. However, while more involved, other methods may be required to confirm the impact of variant loci in sorghum amylase productivity, because many genes cannot simply be transferred between organisms diverged by millions of years of evolution and function in unfamiliar genetic backgrounds. To overcome this a knock-out, a knock-down and knock-in procedure can be performed in-planta.

A knock-out can confirm the function of a gene in-planta succinctly. Once the gene of interest has been identified, as discussed, this knock-out can be created using

CRISPR/Cas9, conveyed by agrobacterium. Once plants with knock-outs events are obtained, the plants will be selfed and progeny with seeds homozygous for the transgene can be malted and assayed. In support of our hypothesis, the results of this experiment would show a reduction in amylase levels in the transformant plants. However, many

61 sorghum accessions show poor regeneration potential that is essential for this and other gene confirmation experiments. Many accessions that do show good regenerability do not show significant amylase productivity. Fortunately, the feasibility of this experiment has been established through studies evaluating the in-vitro regenerability of several sorghum bioenergy lines, including the Grassl genotype (Flinn et al., 2019) Thus, the more easily managed SC 1057 can most be used in this and subsequent in-vitro experiments.

However, a knock-out alone is unlikely to be conclusive, as it may be lethal, halting germination by starving the developing embryo. Therefore, a knock-down experiment using RNAi (RNA interference) constructs can be done in parallel with knock-outs experiments in SC 1057. RNAi, in essence, relies on the expression of RNA that is antisense to the RNA of the gene of interest. This antisense RNA binds to the sense mRNA, forming dsRNA, which molecular machinery in cells identify and cleave

(Wilson & Doudna, 2013). The RNAi used in this experiment should incorporate mutations in the antisense sequence to reduce the binding efficiency to the sense RNA.

This reduction in binding efficiency will result in a “leaky” expression of the original gene, and thus, a knockdown, rather than elimination of expression. Additionally, there are promoters that can drive the production of the RNAi construct, and incompletely halt expression, which can be used (Ecker & Davis, 1986). These RNAi constructs can be introduced through agrobacterium as used in the proposed knock-out pipeline. Plants homozygous for the transgene can have their RNA extracted and malted. Subsequent quantitative-PCR (polymerase chain reaction) using this RNA as a template can confirm the functional knockdown of RNA expression levels for the putative gene. The transgenic

62 malted materials amylase content will also be evaluated and compared to the original untransformed material. If lower levels of amylase are observed in the transgenic material, the impact of this target locus on amylase productivity can be validated. Knock- down methods can be less quantitative than other methods, however this pipeline can reveal loci responsible for SC1057 and other accessions’ high levels of amylase without potential lethality.

Ultimately, the most telling experiment of a gene’s function is to perform a knock-in of the transgene to a surrogate background that does not display the phenotype of interest. If this is possible, it not only confirms the function of the gene, it also indicates this effect can be conferred to others of the same species and closely related organisms. As discussed, a successful knock-in of a productive alpha amylase associated regulator has applications for bioenergy.

Thus, to confirm the target locus’ effect, a knock-in into a low alpha or beta amylase producing line should be performed. Tx430 is a commonly used sorghum in transformations, and much like Grassl, shows appreciable regenerability (Belide et al.,

2017). It also produces low levels of either amylase as observed in our study. Using the same techniques as above to knock-in the SC1057 gene variant into Tx430, it should be possible to determine if the variant gene is indeed responsible for the elevated amylase levels.

Each experiment provides unique insights into the gene’s function that the others cannot. In labs that have established infrastructure to perform these experiments, costs are minimal to validate gene function. Ultimately, determining the causal genetic variation

63 responsible for amylase productivity may improve not only beverage production, but bioenergy production as well.

Alternatives for a potentially complex trait

Amylase productivity appears to be quantitatively inherited based on our observations. There are many biochemical compounds which sorghum produces that have been shown or may inhibit amylase activity. These compounds include apigenins, tannins, and anthocyanins, which are each produced in variable levels within sorghum populations and are regulated by their own set of genes (Mkandawire et al. 2013;

Hargrove et al. 2011). Additionally, amylase expression must be regulated to provide the proper amount of energy to the developing plant throughout germination (Pagano et al.,

1997). Too much expression may starve the plant late in development or provide a nutrient rich environment that increases pathogen susceptibility. Therefore, breeders must account for the potential that both beta and alpha amylase content are complex traits, in their efforts to improve it.

Traditional breeding efforts rely on choices made by phenotypic characteristics to determine the best parents to use in crosses and which progeny to select. An alternative is to use genomically informed choices, which have been made possible due to the ongoing reduction in costs associated with obtaining genomic information. These techniques are called genomic selection and prediction. Summarizing the process employed in genomic selection, markers are first used to determine the genotype of each organism at each QTL.

Then, the effects on the trait of each QTL genotype are estimated. Finally, a sum of all

64 the QTL effects for selection candidates are calculated to obtain their estimated breeding value (Goddard & Hayes, 2007). When this method is used to predict the breeding value of potential offspring from a potential cross, the method is appropriately called genomic prediction.

Genomic selection/prediction informs optimal parental choices and improves subsequent selections. This method aids greatly when applied for the improvement of complex traits, however, may not be as useful for traits with simple inheritance. Genomic prediction and selection inform breeding decisions by leveraging high density whole genome molecular markers that are in linkage disequilibrium with QTLs for the traits of interest (Meuwissen et al., 2001; Goddard & Hayes, 2007). Breeders can also use the technique to inform early selection decisions of material, reducing overall costs. It also can reduce costs by allowing breeders to phenotype a portion of the offspring from a cross, and using genomic prediction, predict the phenotype for the remainder of the genotyped progeny (Crossa et. al., 2017, Myles et al. 2009). In addition to reducing costs, the progeny of these genomics informed breeding typically show greater breeding values

(or their predicted usefulness in making genetic gains for the trait) than those made without such information (Heffner et al, 2010; Zhong et al. 2009).

While initially intensively employed for the improvement of livestock breeding

(Goddard & Hayes, 2009; Meuwissen et. al., 2001; VanRaden et al. 2008), genomic prediction has also been utilized in crops including barley for malting quality (Schmidt et al., 2016; Bhatta et al., 2020), maize for agronomic characteristics (Crossa et al, 2013), wheat for fusarium resistance (Rutkoski et al. 2012), grain yield components and grain

65 composition in sorghum (Sapkota et al., 2020a,b), and could be used for the improvement of sorghum amylase content and complex malting traits.Furthermore, genomic prediction could be employed to great effect in malting sorghum improvement as amylase productivity is a time and cost intensive trait to phenotype.

Summary and conclusions

Sorghum is a diverse species with many uses including beverage making. As a major constituent in the world's most produced liquor (baijiu), and with a growing market in other beverages such as beer, sorghum’s potential for malting and brewing is understudied. While some research has been performed, studies have evaluated samplings that are insufficient to represent the capacity of sorghum with respect to malting quality.

Furthermore, existing collections of sorghum diversity stand to be improved with further incorporation of accessions that have been historically used for these purposes.

Ultimately, the next steps for the improvement of sorghum for beverage making is the identification of markers associated with loci impacting malting characteristics (alpha and beta amylases) through genetic mapping. Two potential methods for genetic mapping of such loci have been discussed. The bioenergy NAM population is positioned to potentially allow for the mapping of loci associated with elevated alpha amylase content without the development of a novel population. The results of mapping efforts will allow breeders to make informed decisions to develop improved malting cultivars. Furthermore, identification of associated loci will facilitate molecular genetic studies that may discover causal variation for the phenotype.

66 Beyond beer and liquor production, understanding the causal genetic variation for amylase content in sorghum could have large implications for the bioenergy market.

Several techniques have been proposed here that would enable verification of any putative genes found by future mapping efforts. Ultimately, such mapping and gene validation efforts may inform the conclusion that amylase content is a complex trait. In this case, there are tools breeders can use in future efforts, including genomic selection and prediction, to aid in the development of superior malting sorghums.

References

Adefila, O. A., Bakare, M. K., & Adewale, I. O. (2012). Characterization of an α-amylase

from sorghum (Sorghum bicolor) obtained under optimized conditions. Journal of

the Institute of Brewing, 118(1), 63–69. https://doi.org/10.1002/jib.11

Aisien, A. O., & Palmer, G. H. (1983). The sorghum embryo in relation to the hydrolysis

of the endosperm during germination and seedling growth. J. Sci. Food Agric.,

34(2), 113–121. doi: 10.1002/jsfa.2740340202

Agu, R. C., & Palmer, G. H. (1997). The effect of temperature on the modification of

sorghum and barley during malting. Process Biochem., 32(6), 501–507. doi:

10.1016/S0032- 9592(97)00002-2

Belide, S., Vanhercke, T., Petrie, J. R., & Singh, S. P. (2017). Robust genetic

transformation of sorghum (Sorghum bicolor L.) using differentiating

embryogenic callus induced from immature embryos. Plant Methods, 13(1), 109.

https://doi.org/10.1186/s13007-017-0260- 9

67 Bhatta, M., Gutierrez, L., Cammarota, L., Cardozo, F., Germán, S., Gómez-Guerrero, B.,

...Castro, A. J. (2020). Multi-trait genomic prediction model increased the

predictive ability for agronomic and malting quality traits in barley (Hordeum

vulgare L.). G3: Genes, Genomes, Genet., 10(3), 1113–1124.

https://10.1534/g3.119.400968

Bouchet, S., Olatoye, M. O., Marla, S. R., Perumal, R., Tesso, T., Yu, J., Tuinstra, M., &

Morris, G. P. (2017). Increased power to dissect adaptive traits in global sorghum

diversity using a nested association mapping population. Genetics, 206(2), 573–

585. https://doi.org/10.1534/genetics.116.198499

Boyles, R. E., Pfeiffer, B. K., Cooper, E. A., Rauh, B. L., Zielinski, K. J., Myers, M. T.,

Brenton, Z., Rooney, W. L., & Kresovich, S. (2017). Genetic dissection of

sorghum grain quality traits using diverse and segregating populations.

Theoretical and Applied Genetics, 130(4), 697–716.

Boyles, R. E., Pfieffer, B. K., Cooper, E. A., Zielinski, K. J., Myers, M. T., Rooney, W.

L., & Kresovich, S. (2017). Quantitative trait loci mapping of agronomic and

yield traits in two grain sorghum biparental families. Crop Science, 57(5), 2443–

2456. https://doi.org/10.2135/cropsci2016.12.0988

Brenton, Z. W., Cooper, E. A., Myers, M. T., Boyles, R. E., Shakoor, N., Zielinski, K. J.,

Rauh, B. L., Bridges, W. C., Morris, G. P., & Kresovich, S. (2016). A genomic

resource for the development, improvement, and exploitation of sorghum for

bioenergy. Genetics, 204(1), 21–33. https://doi.org/10.1534/genetics.115.183947

68 Crossa, J., Beyene, Y., Kassa, S., Pérez, P., Hickey, J. M., Chen, C., ...Babu, R. (2013).

Genomic prediction in maize breeding populations with genotyping-by-

sequencing. G3: Genes, Genomes, Genet., 3(11), 1903–1926.

https://10.1534/g3.113.008227

Crossa, J., Pérez-Rodríguez, P., Cuevas, J., Montesinos-López, O., Jarquín, D., de los

Campos, G., ...Varshney, R. K. (2017). Genomic selection in plant breeding:

methods, models, and perspectives. Trends Plant Sci., 22(11), 961–975. doi:

10.1016/j.tplants.2017.08.011

Ecker, J. R., & Davis, R. W. (1986). Inhibition of gene expression in plant cells by

expression of antisense RNA. Proceedings of the National Academy of Sciences,

83(15), 5372–5376. https://doi.org/10.1073/pnas.83.15.5372

Eichler, J. (2001). Biotechnological uses of archaeal extremozymes. Biotechnology

Advances, 19(4), 261–278. https://doi.org/10.1016/S0734-9750(01)00061-1

Flinn, B. S., Dale, S., Disharoon, A., & Kresovich, S. (2019). Comparative analysis of in

vitro responses and regeneration between diverse bioenergy sorghum genotypes.

BioRxiv, 861328. https://doi.org/10.1101/861328

Frison, C., López, F., & Esquinas-Alcázar, J. (Eds.). (2011). Plant Genetic Resources and

Food Security. FAO, Bioversity International, and Earthscan.

http://www.fao.org/3/a- bb143e.pdf

69 Funnell, D. L., & Pedersen, J. F. (2006). Association of plant color and pericarp color

with colonization of grain by members of Fusarium and Alternaria in near-

isogenic sorghum lines. Plant disease, 90(4), 411-418.

https://doi.org/10.1094/PD-90-0411

Goddard, M. E., & Hayes, B. J. (2009). Mapping genes for complex traits in domestic

animals and their use in breeding programmes. Nature Reviews Genetics, 10(6),

381–391. https://doi.org/10.1038/nrg2575

Goddard, M. E., & Hayes, B. J. (2007). Genomic selection. J. Anim. Breed. Genet.,

124(6), 323– 330. https://10.1111/j.1439-0388.2007.00702.x

Hargrove, J. L., Greenspan, P., Hartle, D. K., & Dowd, C. (2011). Inhibition of aromatase

and α-amylase by flavonoids and proanthocyanidins from Sorghum bicolor bran

extracts. Mary Ann Liebert, Inc., publishers. doi: 10.1089/jmf.2010.0143

Heffner, E. L., Lorenz, A. J., Jannink, J.-L., & Sorrells, M. E. (2010). Plant breeding with

genomic selection: Gain per unit time and cost. Crop Sci., 50(5), 1681–1690. doi:

10.2135/cropsci2009.11.0662

Kumar, R., Singh, S., & Rao, A. (2005). Thermal stability of α-amylase from malted

jowar (Sorghum bicolor). Journal of Agricultural and Food Chemistry, 53, 6883–

6888. https://doi.org/10.1021/jf0501701

Lightbourne, M. (2016). Food Security, Biological Diversity and Intellectual Property

Rights. Routledge.

70 McMullen, M. D., Kresovich, S., Villeda, H. S., Bradbury, P., Li, H., Sun, Q., ...Buckler,

E. S. (2009). Genetic properties of the maize nested association mapping

population. Science, 325(5941), 737–740. https://10.1126/science.1174320

Meuwissen, T. H., Hayes, B. J., & Goddard, M. E. (2001). Prediction of total genetic

value using genome-wide dense marker maps. Genetics, 157(4), 1819–1829.

Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/11290733

Murty, D. S., Nwasike, C. C., & Bello, S. A. (1998). Registration of ‘ICSV 400’ sorghum

cultivar. Crop Sci., 38(6), cropsci1998.0011183X003800060060x–1718.

https://10.2135/cropsci1998.0011183X003800060060x

Pagano, E. A., Benech-Arnold, R. L., Wawrzkiewicz, M., & Steinbach, H. S. (1997).

α-amylase activity in developing sorghum caryopses from sprouting resistant and

susceptible varieties. The role of ABA and GAs on its regulation. Annals of

Botany, 79(1), 13–17. https://doi.org/10.1006/anbo.1996.0296

Phuong, N., Afolayan, G., Stützel, H., Uptmoor, R., & El-Soda, M. (2019). Unraveling

the genetic complexity underlying sorghum response to water availability. PLOS

ONE, 14(4), e0215515. https://doi.org/10.1371/journal.pone.0215515

Plucknett, D. L., Smith, N. J. H., Williams, J. T., & Anishetty, N. M. (1987). Gene Banks

and the World’s Food. Princeton University Press. https://10.2307/j.ctt7ztt0n

Pollard, D. A. (2012). Design and construction of recombinant inbred lines. In S. A.

Rifkin (Ed.), Quantitative Trait Loci (QTL): Methods and Protocols (pp. 31–39).

Humana Press. https://doi.org/10.1007/978-1-61779-785-9_3

71 Rutkoski, J., Benson, J., Jia, Y., Brown-Guedira, G., Jannink, J.-L., & Sorrells, M.

(2012). Evaluation of genomic prediction methods for fusarium head blight

resistance in wheat. Plant Genome, 5(2), 51–61.

https://10.3835/plantgenome2012.02.0001

Sapkota, S., Boyles, R., Cooper, E., Brenton, Z., Myers, M., & Kresovich, S. (2020).

Impact of sorghum racial structure and diversity on genomic prediction of grain

yield components. Crop Science, 60(1), 132–148.

https://doi.org/10.1002/csc2.20060

Sapkota, S., Boatwright, J. L., Jordan, K., Boyles, R., & Kresovich, S. (2020). Multi-trait

regressor stacking increased genomic prediction accuracy of sorghum grain

composition. bioRxiv, 2020.04.03.023531. Retrieved from

http://biorxiv.org/content/early/2020/04/03/2020.04.03.023531.abstract

Schmidt, M., Kollers, S., Maasberg-Prelle, A., Großer, J., Schinkel, B., Tomerius, A.,

...Korzun, V. (2016). Prediction of malting quality traits in barley based on

genome-wide marker data to assess the potential of genomic selection. Theor.

Appl. Genet., 129(2), 203–213. https://10.1007/s00122-015-2639-1

Stevens, C. J., Murphy, C., Roberts, R., Lucas, L., Silva, F., & Fuller, D. Q. (2016).

Between China and South Asia: A Middle Asian corridor of crop dispersal and

agricultural innovation in the Bronze Age. Holocene, 26(10), 1541–1555.

https://doi.org/10.1177/0959683616650268

72 Taylor, J. R. N., Dlamini, B. C., & Kruger, J. (2013). 125th Anniversary Review: The

science of the tropical cereals sorghum, maize and rice in relation to lager beer

brewing. Journal of the Institute of Brewing, 119(1–2), 1–14.

https://doi.org/10.1002/jib.68

Taylor, J. R. N., & Robbins, D. J. (1993). Factors influencing beta-amylase activity in

sorghum malt. J. Inst. Brew., 99(5), 413–416. https://10.1002/j.2050-

0416.1993.tb01181.x

VanRaden, P. M., Van Tassell, C. P., Wiggans, G. R., Sonstegard, T. S., Schnabel, R. D.,

Taylor, J. F., & Schenkel, F. S. (2009). Invited review: Reliability of genomic

predictions for North American Holstein bulls. J. Dairy Sci., 92(1), 16–24.

https://10.3168/jds.2008- 1514

Winchell, F., Brass, M., Manzo, A., Beldados, A., Perna, V., Murphy, C., ...Fuller, D. Q.

(2018). On the origins and dissemination of domesticated sorghum and pearl

millet across Africa and into : A view from the Butana group of the far

eastern Sahel. Afr. Archaeol. Rev., 35(4), 483–505.

https://doi.org/10.1007/s10437-018-9314-2

Wilson, R. C., & Doudna, J. A. (2013). Molecular mechanisms of RNA interference.

Annual Review of Biophysics, 42(1), 217–239. https://doi.org/10.1146/annurev-

biophys-083012-130404

73 Yan, S., Wang, L., Zhao, L., Wang, H., & Wang, D. (2018). Evaluation of genetic

variation among sorghum varieties from southwest China via genome

resequencing. Plant Genome, 11(3), 170098–10.

https://10.3835/plantgenome2017.11.0098

Zhang, Q., & Li, C. (2017). Comparisons of copy number, genomic structure, and

conserved motifs for α-amylase genes from barley, rice, and wheat. Front. Plant

Sci., 8. https://10.3389/fpls.2017.01727

Zheng, Y., Wood, D., Wang, H. H., & Jones, J. P. H. (2018). Predicting potential impacts

of China’s retaliatory tariffs on the U.S. farm sector. Choices, 33(2), 1–6. JSTOR.

74

APPENDICES

75 Appendix A

FIGURES

Figure A-1: Mean (a) alpha and (b) beta amylase levels observed in samples as compared to plant introgression number. Texture in the bars represent the pericarp color of the accession. Error bars represent one standard deviation.

76

Figure A-2: Geographic distribution of materials sampled originating in Africa and their classification into regions.

77

Figure A-3: (a) Alpha and (b) beta amylase content between pericarp colors of sorghums show significant differences.

78

Figure A-4: Alpha amylase versus (a) tannins and (b) phenols. Beta amylase versus (c) tannins and (d) phenols. Formula of the linear model and R2 of the model is given.

79

Figure A-5: Geographic distribution of material surveyed in this study.

80 Appendix B

TABLES

Plant ID Plant Name Origin PI 282837 Kouran N'Da' PI 282838 Kouran Kass Chad PI 482613 Chibuka Zimbabwe PI 482641 TGR 195 Zimbabwe PI 482662 Mutode Zimbabwe PI 482731 Chikota Zimbabwe PI 482736 TGR 505 Zimbabwe PI 482738 Gwengwere Zimbabwe PI 482826 Rundende Zimbabwe PI 482845 Bumbi Zimbabwe PI 482846 Musoso Zimbabwe PI 482847 TGR 1347C Zimbabwe PI 482848 TGR 1347E Zimbabwe PI 482948 Chisiri Zimbabwe PI 494913 ZFA 3386 Zambia PI 506010 Naga Togo PI 506011 Yarba Togo PI 506012 Wilnaga Togo PI 506013 Belko Togo PI 506018 Dimoni Togo PI 506019 Dimoni Togo PI 506020 Tchali Togo PI 506021 Tchali Togo PI 506022 Tchali Togo

81 PI 506023 Tchelouol Togo PI 506024 Tchalpieni Togo PI 506027 Outchitawa Togo PI 506029 Kpana Togo PI 506030 Tchinlouol Togo PI 506039 Belko Togo PI 506042 Itchalimon Togo PI 506047 Gbesho Yali Togo Djafougou- PI 506059 Togo Hondidjona PI 506094 Idipin Togo PI 508248 4184 PI 524516 Hargha PI 524521 Kareng Sudan PI 524538 Mareg Sudan PI 524539 Feterita Sudan PI 524540 Dawa Sudan PI 524557 Mareg Ahmar Sudan Mareg Kubar PI 524565 Sudan Aswad PI 524599 Maramaka Sudan PI 524600 Maramaka (red) Sudan PI 524601 Ariana Sudan PI 524602 Karamaka (white) Sudan PI 524603 Karamaka (red) Sudan PI 524604 Wee Sudan PI 524605 IBPGR 37 Sudan PI 524607 Kulum Sudan

82 PI 524608 Kulum Sudan

PI 524609 Bakhit Sudan PI 524610 Bakhit Sudan PI 524611 Bakhit Sudan PI 524647 Ghala Baida Sudan PI 524648 Najjad Sudan PI 524649 Abu Hamir Sudan PI 524651 Chokoria Sudan PI 524652 85 Sudan PI 524654 Huria Sudan PI 524656 Malak Sudan PI 524657 Chokoria Sudan PI 524658 Kurgi Sudan PI 524659 Saliy Sudan PI 524660 Um Kurundu Sudan PI 524661 Farik Sudan PI 524662 Baagosh Sudan PI 524663 96 Sudan PI 524664 Ghala Kabish Sudan PI 524665 Mashrula Sudan PI 524668 Kulum Sudan PI 524669 Huria Sudan PI 524670 Malek Sudan PI 524673 Huria Sudan PI 524674 Nyily-Nyily Sudan PI 524675 107 Sudan PI 524676 Sefiera Sudan

83 PI 524677 Robai Sudan PI 524678 Kurgi Sudan PI 524679 Kulum Sudan

PI 524680 Barky Sudan

PI 524681 Nyily-Nyily Sudan PI 524682 Gasaby Sudan PI 524684 Tobala Sudan PI 524685 Chokoria Sudan PI 524686 Najjad Sudan PI 524687 Um Shaishilate Sudan PI 524688 Dobar Sudan PI 524691 Kabai-Kabai Sudan PI 524693 Kulum Sudan PI 524694 Mayo Sudan PI 524695 Wad Margani Sudan PI 524696 Mugud Sudan PI 524697 Habashiya Sudan PI 524699 Tapira Sudan PI 524700 Tapira Sudan PI 524702 Tapira Sudan PI 524704 Tapira Beida (white) Sudan PI 524705 Feterita Sudan PI 524706 138 Sudan PI 524707 Zinnary Sudan PI 524717 Zinnary Sudan

Table B-1: Beer sorghum germplasm collected by the USDA GRIN (https://npgsweb.ars- grin.gov).

84

Table B-2: Data used in this study.

85 Male Bioenergy NAM Parents

Alpha-Amylase Beta-Amylase Taxa ID Common CU/g U/g 22913 Chinese 224.1 6.1 Amber 229841 IS 2382 164.8 1.3 297130 IS 13613 249.7 6.2 297155 IS 13633 131.1 13.5 329311 IS 11069 237.8 6.7 506069 Mbonou 85.5 1.3 508366 MA 38 222.2 5.4 510757 AP79-714 86.1 1.7 586454 Leoti 66.2 9.5 651496 Rio 39.3 6.0 655972 Pink Kafir 119.2 0.6 Control Barley 274.0 19.2 Control

Table B-3: Amylase content of bioenergy NAM parents.

86