Selection for high β-glucan content and good agronomic performance in oat

grain

Alona A. Chernyshova

Professor: Jean-Luc Jannink

Iowa State University

Department of Agronomy

American Oat Workers Conference

Fargo, ND

25 July 2006

2

Outline

♦ Introduction

♦ Objectives

♦ Materials and methods

♦ Results

♦ Conclusions

3

♦ -glucan(1-3)(1-4)-β-D is a soluble fiber found in the endosperm and subaleurone layers of cereal seeds

Introduction

4

Introduction

The structure of oat groat

♦ -glucan(1-3)(1-4)-β-D is a soluble fiber found in the endosperm and subaleurone layers of cereal seeds

-glucan is responsible for many health related effects of oat:

• reduction of blood cholesterol levels

• reduction of the risk of coronary heart disease

• reduction of the risk of diabetes

• attenuation of blood glucose and insulin, and

• prolonged satiety

Standard oat varieties contain from 4.5 to 5.0 % of -glucan

We would like to double this concentration

5

Introduction

♦ Viscosity is the thickness or resistance to flow of a liquid

Many health related effects of oats result from the high viscosity of their soluble β-glucans

Regression analyses have shown a positive effect of -glucan content on viscosity

♦ Viscosity deviation

Deviation of viscosity from its prediction based on -glucan content

We call this the “viscosity deviation”

Viscosity of some oat lines deviated significantly from the regression prediction, indicating that, for a given -glucan content, some lines generate more viscosity and some less viscosity

6

Objectives

♦ to estimate the genetic components of variance in -glucan content, viscosity, and viscosity deviation in high -glucan lines, elite agronomic lines, and in their population crosses

♦ to evaluate the differences between elite agronomic lines and high -glucan lines for -glucan, viscosity, and viscosity deviation

♦ to use a powerful population design to detect epistatic interaction among parents

7

Materials and methods

Figure 2. Experimental program

F1 generation

represented by 48 crosses

F3 individuals derived

from F1 by selfing

In 2005 lines were planted in two locations with four replications and analyzed for beta-glucan

and viscosity

Twelve lines selected for high beta-glucan content

from the Iowa State University Oat Breeding

Program

Twelve lines selected for good agronomic

performance from the Uniform Oat

Performance Nursery coordinated by the

USDA

8

Materials and methods

Beta-glucan parents

1 2 3 4 5 6 7 8 9 10 11 12

1 * * * *

2 * * * *

3 * * * *

4 * * * *

5 * * * *

6 * * * *

7 * * * *

8

* * * *

9 * * * *

10 * * * *

11 * * * *

Agr

onom

ic p

aren

ts

12

* * * *

North Carolina Design II

Mating design

9

Materials and methods

♦ The -glucan concentration in flours was determined enzymatically using a mixed linkage -glucan kit (Megazyme Int., Wicklow, Ireland)

♦ Modifications were made to the Approved Method, which allowed us to increase the number of samples analyzed per unit time and cost spent

Micro Enzymatic Method

10

Materials and methods

Micro Enzymatic Method

11

Materials and methods

Microplate reader measures the light absorbance in each cell that reflects the amount of

β-glucan in the sample

Micro Enzymatic Method

12

Materials and methods

Glucose standards

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 100 200 300 400 500 600 700

Glucose standards, µg/mL

Op

tical

den

sity

13

Materials and methods

Glucose standards

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 100 200 300 400 500 600 700

Glucose standards, µg/mL

Op

tica

l den

sity

14

Materials and methods

♦ Viscosity was measured using a Rapid Visco Analyser (RVA, Newport Scientific Warriewood, Australia), as a function of temperature, time, and stirring

Viscosity Measurements

15

Materials and methods

The Rapid Visco-Analyser is used to measure the apparent viscosity of samples as a function of temperature, time, and stirring

16

Small Preliminary Experiment

♦ to maximize the solubilization of -glucan over a short period of time while also avoiding a contribution of starch gelatinization to viscosity

♦ Jim, Paul, IA95111, and N979-5-2-4

♦ -amylase was added to eliminate the contribution of starch to the viscosity of oat flour

♦ Constant temperature 45 °C or constant temperature 55 °C

17

♦ Data was analysed with SAS proc GLM

♦ All factors (oat line, temperature, and presence / absence of -amylase, and all pair-wise interactions) were considered fixed

♦ The dependent variables were log-transformed peak viscosity and log-transformed average viscosity between 5 to 7 minutes

Small Preliminary Experiment

18

♦ No effect of -amylase

♦ At these temperatures starch does not contribute to viscosity

♦ Oat line (P < 0.001)

♦ Mean log-transformed peak viscosity Jim, Paul, IA95111, and N979-5-2-4 were 4.25, 4.42, 5.05, and 5.69, respectively

♦ The effect of temperature was marginally significant (P = 0.08)

♦ Mean log-transformed peak viscosity of 4.81 and 4.89 for 40 °C and 55 °C, respectively

♦ Correlation (r2=0.99) between log-transformed peak viscosity and log-transformed average viscosity between 5 to 7 minutes

Small Preliminary Experiment

19

♦ Silver nitrate inactivates -glucanases present in the flour and originating from grain or microbial contamination

♦ Log peak viscosities with and without silver nitrate were 4.36 versus 4.36 for Jim, and 5.12 versus 5.07 for IA95111

♦ We chose to use deionized water for all subsequent experiments

Small Preliminary Experiment

20

0

100

200

300

400

500

600

700

800

900

0 1 2 3 4 5 6 7 8 9 10

Time, min

Vis

cosi

ty,

RV

U

IA03146-6 (1) IA03146-4 (2) IA03162-4 (3)

IA03196-6 (4) IA03146-7 (5)

Materials and methods

Viscosity Measurements

The viscosity measurement was the average viscosity from 5 to 7 minutes on the pasting curve

21

Nutritional function and good agronomic performance in oat

Statistical analysis Beta-glucan parents

1 2 3 4 5 6 7 8 9 10 11 12

1 * * * *

2 * * * *

3 * * * *

4 * * * *

5 * * * *

6 * * * *

7 * * * *

8

* * * *

9 * * * *

10 * * * *

11 * * * *

Agr

onom

ic p

aren

ts

12

* * * *

North Carolina Design II

♦ Lines were planted in an incomplete block design with four replications at each of two environments (Ames and Kanawha)

♦ A set, which consisted of four inbred agronomic parents, four inbred -glucan parents, and 16 F3:4 progeny

lines, was assigned to each incomplete block

♦ In each replication, a family was represented by a different F3:4

progeny line

22

Nutritional function and good agronomic performance in oat

Statistical Analysis

3.5

4

4.5

5

5.5

6

6.5

30 35 40 45 50 55 60 65 70 75 80

Beta-Glucan Content g kg-1

Log(

RV

A),

Log

[RV

U]

Progeny Parents regression line

♦ The relationship between flour -glucan content and flour slurry viscosity was best linearized by taking the logarithm of slurry viscosity

r2=0.35

23

♦ All traits showed significant genetic variance

♦ Viscosity and grain yield showed significant specific combining ability variance indicating interactions between loci affecting these traits

Nutritional function and good agronomic performance in oat

Results and discussion

24

Nutritional function and good agronomic performance in oat

Results and discussion

H2 = [var(Fam) + var(L)] / [var(Fam) + var(L) + var(Err)]

♦ These heritabilities were high, ranging from 0.42 for the viscosity deviation to 0.82 for viscosity itself

25

Conclusion

♦ The study found relatively high heritabilities for the traits -glucan content, flour slurry viscosity, and viscosity deviation that are relevant for the selection of oat lines with high nutritional function

♦ The study confirmed that the mode of gene action of loci affecting -glucan was primarily additive, but found observable interallelic interaction variance for flour slurry viscosity

♦ This interaction variance may further explain the transgressive segregation for flour slurry viscosity that was observed

26

Conclusion

♦ Finally, the study showed for the first time that selection for viscosity deviation is feasible

♦ While at this time we do not know what are the mechanisms generating viscosity deviations, the fact that the trait is selectable should allow us to divergently select for the trait and to study the differences among lines with high and low viscosity deviation

27

Acknowledgements

♦ My major professor:

Jean-Luc Jannink

♦ My committee members:

Pamela J. White

M. Paul Scott

♦ Dr. Sedat Sayar and Dr. Ni Yao

♦ Merinda Struthers

♦ George Patrick and Ron Skrdla

♦ All graduate students in my group

♦ USDA-NRI Competitive Grants Program, award number 2004-02413

Thank you

28

Nutritional function and good agronomic performance in oat

♦ Variance components and predictors of line effects were estimated in a Bayesian analysis using WinBUGS

Pijkm = Blki + Popj + Parentk(j) + Errorm

Liabcd = Blki + *iabcd + + Agra + Betb + Agr*Betab + Linec(ab) + Errord

Pijkm = Blki + *ijkm + Popj + Parentk(j) + Errorm

Liabcd = Blki + + Agra + Betb + Agr*Betab + Linec(ab) + Errord

Statistical Analysis

29

Nutritional function and good agronomic performance in oat

Results and discussion

2GCA_A 2

GCA_B 2SCA var(Fam) var(L) var(Err)

-glucan g kg-1 0.058NS 0.064NS 0.103NS 0.225* 0.485* 0.353*

Log RVA log(RVU) 0.030NS 0.016NS 0.030* 0.075* 0.131* 0.044*

Visc. Dev. log(RVU) 0.018* 0.015* 0.012NS 0.045* 0.016NS 0.084*

Grain yield

g m-2 2400* 3900* 9900* 17000* 4500* 10676*

30

Nutritional function and good agronomic performance in oat

Results and discussion

Population -glucan content Log RVA Grain yield

  g kg-1 Log (RVU) g m-2

Agronomic 0.75 * 0.82 * -0.08NS

-glucan 0.03NS 0.66 * 0.87 *

*P < 0.05