Transcript Document

Research Experience in Molecular Biotechnology & Genomics
Summer 2008
Center for Integrated Animal Genomics
Doyle, James E. Jr, Ruth A. Swanson-Wagner,1,2 Xuefeng Zhao,3,4 Dan Nettleton,5 and Patrick S. Schnable1,2,4
1. Interdepartmental Genetics Graduate Program and Department of Agronomy, 2. Department of Genetics, Development and Cell Biology, 3. Laurence H. Baker Center for Bioinformatics and Biological Statistics, 4. Center for Plant Genomics, 5. Department of Statistics
Global Profiling in Maize Reciprocal Hybrids Reveals Extensive Differential Gene
Expression
Introduction and Summary
Results and Discussion
Heterosis is the phenomenon whereby F1 offspring of two different inbred
parents exhibit phenotypic characteristics superior to parental levels. Heterotic
traits, such as yield, stress tolerance, root growth, and height have been widely
exploited by modern agriculture. A classic example of heterosis is the B73 and
Mo17 inbred lines which produce a heterotic F1 hybrid and have been the
subjects of numerous recent studies (1-6). The Intermated B73 and Mo17
Recombinant Inbred Lines (IBM RILs) derived from the B73xMo17 cross have
been used to produce a genetic map in maize with thousands of markers (7).
Previous studies have detected differential expression of genes between B73,
Mo17, and their hybrid (1-4).
The B73/Mo17 hybrid can be generated in two ways. In one case B73 is
used as the female parent (B73xMo17); in the other case Mo17 is used as the
female (Mo17xB73) (Figure 4). Progeny from these two crosses are termed
“reciprocal hybrids”. Even though they have identical DNA content reciprocal
hybrids often exhibit phenotypic differences (8,9), such as seedling dry weight
(Figure 1; Table 1. Swanson-Wagner et al, unpublished). Hence, it is surprising
that published global profiling studies of B73/Mo17 reciprocal hybrids have
found few differences in gene expression (1,5,6). In contrast, analyses of gene
expression patterns in an expression QTL (eQTL) experiment conducted in the
Schnable lab that compared heterozygous and homozygous genotypes provided
suggestive evidence that reciprocal effects exist in maize (Swanson-Wagner et
al., unpublished). The magnitude of these differences in gene expression
between reciprocals is, however, modest (Figure 2). We therefore hypothesized
that the prior studies (1,5,6) did not have sufficient statistical power to detect
these reciprocal cross differences.
To overcome this limitation with published studies, we used ten biological
replications to test for differential gene expression between two pairs of
reciprocal hybrids (Figure 5). Over 2,000 genes were estimated to be
differentially expressed in each pair of reciprocal hybrids, demonstrating that
reciprocal effects on gene expression are widespread and can be detected in
experiments with sufficient statistical power.
B73xMo17 vs. Mo17xB73
p-values
Fold Changes For Sig. Genes
Log Fold Change
Figure 2. (Left) p-Value histogram for the test of differential expression in
B73xMo17 vs. Mo17 vs. B73. (Right) Histogram of fold changes for significant
genes in B73xMo17 vs. Mo17xB73 (Swanson-Wagner et al, unpublished).
B73x
B73x M0024
M0024 M0024 xB73
M0024 M0024 B73x
xB73
xB73 M0024
M0024 M0024 B73x
xB73
xB73 M0024
Ten biological replications of B73xM0024 vs. M0024xB73 and ten replications of
Mo17xM0024 vs. M0024xMo17 were hybridized to the SAM 1.2 microarray. The
14,118 informative spots were filtered and analyzed using custom R scripts,
providing a p-value for each spot for the test of whether it is differentially
expressed in the reciprocal hybrids (Figure 7). Over 2,000 genes were estimated to
be differentially expressed in each pair of reciprocal hybrids (12) (Table 2).
B73
Reciprocal
Hybrids
Mo17
Figure 1. Inbred parents and reciprocal hybrids (14 days old).
Mean Seedling Dry
Genotype
Weight (g)
B73
0.351 ± 0.092
B73xMo17
0.392 ± 0.085
Mo17xB73
0.517 ± 0.078
Mo17
0.295 ± 0.077
N=36 seedlings per genotype
Table 1. Biomass
accumulation in maize
hybrids and their inbred
parents.
Analysis
# Reps
Estimated Number Diff.
Expressed
B73xMo17 vs. Mo17xB73
6
7,338 (51.9%)
B73xM0024 vs. M0024xB73
10
2,351 (16.7%)
Mo17xM0024 vs. M0024xMo17
10
2,318 (16.4%)
Table 2. Number of differentially expressed genes estimated in each reciprocal
pair for a total of 14,118 assayed.
Materials and Methods
Generating Stocks
IBM RIL M0024 was crossed as both male and female to both B73 and Mo17 to
generate four F1 hybrids (RILxB73, B73xRIL, RILxMo17, and Mo17xRIL).
Mo17xM0024 vs. M0024xMo17
B73xM0024 vs. M0024xB73
Experimental Design
Ten biological replications comparing the reciprocal hybrids were grown under highly
controlled conditions in growth chambers (Percival Scientific, Perry, IA) to reduce
environmental variation. Per replication, nine seedlings of each genotype were
randomly assigned to a position within a tray (see Figure 3).
Plant Growth
Plants were grown and harvested in accordance with the methods of Swanson-Wagner
et al (3).
Sample Preparation
RNAs were extracted from ~10g of tissue pooled from six seedlings using homemade
TRIzol (10). mRNAs were then isolated using the OligoTex kit (Qiagen, Valencia, CA).
Two µg of mRNA were reverse transcribed using Superscript II (Invitrogen, Carlsbad,
CA) and labeled (with Cy3 and Cy5) in accordance with the experimental design (Figure
5) (11) . Samples were hybridized to the SAM 1.2 (Figure 6) cDNA microarray (GPL
4521) for 12-18 hours at 42C, washed, and scanned six times each at constant PMT and
increasing laser power with a ProScan Array HT (Perkin Elmer, Wellesley, MA).
Quantification and Analysis
Quantification of slides was completed using Imagene 8.0 software (BioDiscovery, El
Segundo, CA), and a single scan for each slide was picked for analysis using linear
regression models. 14,118 informative spots were normalized in R using Lowess
normalization. The analysis testing for differential expression between the two
genotypes (B73xRIL vs RILxB73 and Mo17xRIL vs RILxMo17) was completed using
custom R scripts.
B73x M0024 M0024
xB73
M0024 xB73
M0024 M0024 B73x
xB73
xB73 M0024
B73x
B73x
B73x
M0024 M0024 M0024
Figure 3. Example plant
randomization. Six plants per
genotype were randomly chosen
to be harvested.
In our prior analysis of the B73xMo17 vs. Mo17xB73 reciprocal hybrids over
7,000 genes were estimated to be differentially expressed (Figure 2). On average
half as many genes should exhibit allelic differences between M0024 and B73 and
between M0024 and Mo17 as compared to between B73 and Mo17 (Figure 4).
This can likely explain the smaller number of genes that exhibit reciprocal cross
differences in the current study (~2,300) as compared to the study involving
B73/Mo17 reciprocal hybrids (~7,300).
p-value
p-value
Figure 7. P-value histograms for 14,118 tests for differential expression in Mo17xM0024
vs. M0024xMo17 and B73xM0024 and M0024xB73. Over 2,000 genes were estimated
to be differentially expressed in both experiments.
Although phenotypic differences in reciprocals have been observed, previous
studies have not reported differences in gene expression. Here, with sufficient
replication we were able to detect differences. The implications of this range from
basic research to applied research. Because the differences in phenotype (whether
on a molecular level or measured in the field) can depend on epigenetics or parentof-origin, genotype alone may not be a stable predictor of behavior for the gene or
phenotype of interest. Nor is it safe to assume that a hybrid will behave similarly
to its reciprocal cross. Thus, integrating traits of interest into crops will require a
detailed understanding of the regulation mechanisms of the trait and the
transmission of the allele. A better understanding of the mechanisms of regulation
can be obtained by mapping the reciprocal differences to genomic loci and
studying the various mechanisms (methylation, parent-of-origin effects, etc.) of
specific genes that exhibit reciprocal effects. Functional annotation of the
significant genes may also provide insight into biological processes or pathways
that may be contributing to the observed phenotypic differences.
References
(1) Stupar, R.M. & Springer, N.M. (2006) Genetics 173, 2199-2210.
Figure 4. An example of the crosses being compared
in this study. The offspring at the bottom make up a
reciprocal pair, and are genomically identical, with
the difference being which parent is the male parent,
and which is female.
Figure 5. Experimental design. Ten
biological replications for each
reciprocal hybrid pair were hybridized
to the chip with randomized dye
assignment. A similar design was used
for the cross between M0024 and Mo17
Figure 6. A portion of the SAM 1.2 microarray chip.
Acknowledgements
We thank Marianne Smith for contributions and support in sample processing and
hybridization, undergraduate lab assistants Eli Berhane, Joe Gebelein, Tom Kemerrer,
Danielle Reed, and Katie Rabovsky for assistance in sample processing, and the PSI
Genomics Technologies Facility for printing the arrays.
Correspondence should be addressed to [email protected].
(2) Auger, D.L., Peters, E.M., & Birchler, J.A. (2005) Journal of Heredity 96, 614-617.
(3) Swanson-Wagner, R.A., Jia, Y., DeCook, R., Borsuk, L.A., Nettleton, D., & Schnable, P.S. (2006) Proc. Nat.
Acad. Sci. USA 103, 6805-6810.
(4) Stupar, R.M., Gardiner, J.M., Oldre, A.G., Haun, W.J., Chandler, V.L., & Springer, N.M. (2008) BMC Plant
Biol. 8, 33.
(5) Guo, M., Rupe, M.A., Danilevskaya, O.N., Yang, X., & Hu, Z. (2003) Plant Journal 36, 30-44.
(6) Frascaroli, E., Cane, M.A., Landi, P., Pea, G., Gianfranceschi, L., Villa, M., Morgante, M., & Enrico Pe, M.
(2007) Genetics 176, 625-644.
(7) Fu, Y., Wen, T.-J., Ronin, Y.I., Hsin, C.D., Guo, L., Mester, D.I., Yang, Y., Lee, M., Korol, A.B., Ashlock,
D.A., Schnable, P.S. (2006) Genetics 174, 1671-1683.
(8) Mann, C.E. & Pollmer, W.G. (1981) Maydica 26, 263-271.
(9) Pollmer, W.G., Klein, D., & Dhillon, B.S. (1979) Euphytica 28, 325-328.
(10) Chomczynski, P. & Sacchi, N. (1987) Anal Biochem 162, 156-159.
(11) Nakazono, M., Qiu, F., Borsuk, L. A. & Schnable, P.S. (2003) Plant Cell 15, 583-596.
(12) Ruppert, D., Nettleton, D., & Hwang, J.T. (2007) Biometrics 63, 483-495.
Program supported by the National Science Foundation Research Experience for Undergraduates
DBI-0552371