A community resource for high-throughput quantitative RT-PCR analysis of transcription factor gene expression in <it>Medicago truncatula</it>

1746-4811-4-18 1746-4811 Methodology A community resource for high-throughput quantitative RT-PCR analysis of transcription factor gene expression in <it>Medicago truncatula</it> Kakar Klementina kakar@mpimp-golm.mpg.de Wandrey Maren wandrey@gfz-potsdam.de Czechowski Tomasz tc519@york.ac.uk Gaertner Tanja gaertner@mpimp-golm.mpg.de Scheible Wolf-Rüdiger scheible@mpimp-golm.mpg.de Stitt Mark stitt@mpimp-golm.mpg.de Torres-Jerez Ivone itjerez@noble.org Xiao Yongli yxiao@jcvi.org Redman C Julia jredman@jcvi.org Wu C Hank hwu@jcvi.org Cheung Foo FCheung@jcvi.org Town D Christopher cdtown@jcvi.org Udvardi K Michael mudvardi@noble.org

Max-Planck Institute of Molecular Plant Physiology, Am Mühlenberg 1, 14476 Potsdam-Golm, Germany

The J. Craig Venter Institute, 9704 Medical Center Drive, Rockville, MD, 20850, USA

The Samuel Roberts Noble Foundation, 2510 Sam Noble Parkway, Ardmore, OK, 73401, USA

Plant Methods 1746-4811 2008 4 1 18 http://www.plantmethods.com/content/4/1/18 18611268 10.1186/1746-4811-4-18 15 4 2008 08 7 2008 08 7 2008 2008 Kakar et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background

Medicago truncatula is a model legume species that is currently the focus of an international genome sequencing effort. Although several different oligonucleotide and cDNA arrays have been produced for genome-wide transcript analysis of this species, intrinsic limitations in the sensitivity of hybridization-based technologies mean that transcripts of genes expressed at low-levels cannot be measured accurately with these tools. Amongst such genes are many encoding transcription factors (TFs), which are arguably the most important class of regulatory proteins. Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) is the most sensitive method currently available for transcript quantification, and one that can be scaled up to analyze transcripts of thousands of genes in parallel. Thus, qRT-PCR is an ideal method to tackle the problem of TF transcript quantification in Medicago and other plants.

Results

We established a bioinformatics pipeline to identify putative TF genes in Medicago truncatula and to design gene-specific oligonucleotide primers for qRT-PCR analysis of TF transcripts. We validated the efficacy and gene-specificity of over 1000 TF primer pairs and utilized these to identify sets of organ-enhanced TF genes that may play important roles in organ development or differentiation in this species. This community resource will be developed further as more genome sequence becomes available, with the ultimate goal of producing validated, gene-specific primers for all Medicago TF genes.

Conclusion

High-throughput qRT-PCR using a 384-well plate format enables rapid, flexible, and sensitive quantification of all predicted Medicago transcription factor mRNAs. This resource has been utilized recently by several groups in Europe, Australia, and the USA, and we expect that it will become the 'gold-standard' for TF transcript profiling in Medicago truncatula.

Background

Legumes are second only to grasses in agricultural importance 1. They are a mainstay of sustainable agricultural systems because of their ability to reduce atmospheric nitrogen (N₂) to ammonia via a symbiosis with bacteria called rhizobia. This provides legumes and subsequent crops with a free and renewable source of nitrogen in lieu of expensive, environmentally-unfriendly fertilizers. Development and differentiation of root nodules, the organ that accommodates nitrogen-fixing rhizobia in legumes, is orchestrated by transcription factors 23456789. Transcription factors are DNA-binding proteins that regulate the transcription of most, if not all genes 10. As a result, TFs play central roles in all aspects of plant biology, including development and differentiation of organs and adaptive responses to changes in the environment 11. Transcription factors as a whole are an important target of plant research because they are a key to understanding the regulation of important plant processes as well as potential tools to optimize these processes for agriculture.

The importance of TFs in plant biology is reflected by the fact that approximately 5% of all plant genes encode such proteins 10. Thus, even species with relatively small genomes, such as Arabidopsis thaliana contain thousands of TF genes 10. This presents a real challenge for systematic approaches to decipher the function of TF genes in plants. Classical, 'forward' genetics has uncovered the roles of perhaps a hundred TF genes in Arabidopsis 12 and far fewer in other species 11. Reverse-genetic approaches, using T-DNA insertion mutants for instance 13, provide a means to decipher in a systematic and relatively rapid manner the function of TF genes/proteins, although gene-redundancy often stymies this enterprise 12. Another stumbling-block is that phenotypes associated with non-redundant TFs may be subtle in nature.

Transcript profiling can help to uncover the functions of TF genes/proteins by revealing where and when in a plant TF genes are expressed. This information can help direct our attention to particular organs, developmental stages, or conditions under which aberrant phenotypes might become apparent in a TF mutant of interest.

Medicago truncatula is a model legume species that is currently the focus of an international genome sequencing effort 14. Several generations of cDNA 15 and oligonucleotide arrays 16 have been developed for transcriptome analysis of Medicago truncatula, including most-recently an Affymetrix GeneChip that contains 51,000 probe-sets representing a large proportion of all the genes in this species 17. While these tools now provide a means to measure the transcriptional output of a large proportion of genes in Medicago, inherent limitations in the sensitivity of hybridization-based technologies 18 mean that transcripts of a substantial number of genes cannot be detected even when probes for these transcripts are present on the array/chip. Furthermore, expansion of arrays to encompass novel genes uncovered by genome sequencing is not a trivial task. An alternative to arrays that is 2–3 orders of magnitude more sensitive and more flexible in terms of expansion to encompass novel genes is quantitative reverse transcription-polymerase chain reaction (qRT-PCR). Platforms for qRT-PCR analysis of thousands of Arabidopsis and rice TF genes have been developed by us and others 1920, and utilized to identify TF genes involved in Arabidopsis responses to nutrient stress and pathogen attack 21222324. Here we describe a bioinformatics pipeline to identify putative TF genes in Medicago truncatula and to design gene-specific oligonucleotide primers for qRT-PCR analysis of all predicted TF transcripts. Over 1000 TF primer pairs were tested and used to identify sets of organ-enhanced TF genes that may play important roles in organ development or differentiation in this species.

Results and Discussion

Identification of putative transcription factors

TF protein families are generally defined by the type(s) of DNA-binding domain they contain and putative TF genes are often identified on the basis of DNA sequences that encode known DNA-binding domains 10112526. We utilized this approach to identify putative TFs of Medicago amongst the set of proteins predicted from genomic sequence by the International Medicago Genome Annotation Group (IMGAG). Proteins of IMGAG release 1, which contained over 40,000 predicted proteins, were screened for the presence of known or presumed DNA-binding domains (Table 1), using INTERPRO 27. Medicago proteins containing putative DNA binding domains and other domains associated with TFs were then used as query sequences in WU-BLASTX 28 which included searches of both the non-redundant DNA database of NCBI 29 and the well-curated protein database, UniProt 30 to check annotations of related proteins in support of tentative Medicago TF assignments. This process resulted in a list of 1045 putative TF genes (see Additional file 1). We utilized genomic sequences rather than the large collection of partial cDNA sequences present in Expressed Sequence Tag (EST) databases for Medicago as the starting point for TF gene discovery because protein sequences derived from genomic sequence are more complete and the set of IMGAG proteins essentially contains no redundancy. Although identification of the 'complete' set of Medicago TFs from IMGAG-annotated proteins will only be possible upon completion of genome sequencing, we expect little or no redundancy in the protein set targeted by our primer set. This approach avoids wasting money on redundant primer sets and re-organization of primers when redundancy is detected, both of which would have been inevitable if we chose to use ESTs in addition to genomic sequences to identify Medicago TFs.

Additional file 1

Complete list of TF genes, primer sequences and corresponding PCR efficiencies. The TF primer platform was established based on the data for the gene models according to SA, specific amplification; NA, no amplification; NS, non-specific amplification. RT-PCR products of primer sequences indicated in bold were sequenced.

Click here for file

Table 1

Classification of putative transcription factors of Medicago into families and sub-families

TF Family

No. of

Genes

Characteristic Domain

(InterPro No.)

Domain

Function

Domain Description

MYB/HD-like

IPR001005, IPR009057

Myb, DNA-binding; homeodomain like

MYB

IPR001005

Myb, DNA-binding

C₂H₂(Zn)

IPR007087

Zn-finger, C2H2 type

AP2/EREBP

IPR001471

Pathogenesis-related transcriptional factor and ethylene response factor

BHLH

IPR001092

Basic helix-loop-helix dimerisation region bHLH

HD-like

IPR009057

Homeodomain like

HD family

IPR001356

Homeobox

HD-ZIP

IPR006712

HD-ZIP protein, N terminus

HD-PHD finger

IPR001965

Zn-finger like, PHD-finger

MADS

IPR002100

TF, MADS-box

BZIP

IPR004827

Basic Leu zipper (bZIP) TF

PHD

IPR001965

Zn-finger like, PHD-finger

WRKY family

IPR003657

DNA-binding WRKY

WRKY

LLR WRKY

IPR001611

Leu-rich repeat

ABI3/VP1

IPR003340

TF B3

NAC

IPR003441

No apical meristem (NAM) protein

C₃H-type1(Zn)

IPR000571

Zn-finger, C-x8-C-x5-C-x3-H type

ARF

IPR003340, IPR010525, IPR011525

JUMONJI

IPR003347

TF jumonji, jmjC

GRAS

IPR005202

GRAS TF

HMG

IPR000637

HMG-I and HMG-Y, DNA binding

AS2

IPR004883

Lateral organ boundaries

C₂C₂(Zn)

Dof

IPR003851

Zn-finger, Dof type

GATA

IPR000679

Zn-finger, GATA type

CO-like

IPR000315

Zn-finger, B-box

YABBY

IPR006780

YABBY protein

CCAAT-HAP3 type

IPR003958

TF CBF/NF-Y/archaeal histone

GRF

IPR010666

Zn-finger, GRF type

SBP

IPR004333

SBP

EIL

IPR006957

Ethylene insensitive 3

LIM

IPR001781

Zn-binding protein, LIM

SNF2

IPR000330

SNF2 family N-terminal domain

E2F/DP

IPR003316

TF E2F/dimerisation partner (TDP)

TCP

IPR005333

TCP TF

FHA

IPR000253

Forkhead-associated

ARID

IPR001606

AT-rich interaction region

HSF

IPR000232

Heat shock factor (HSF)-type, DNA binding

AUX/IAA

IPR003311

AUX/IAA protein

SRS

IPR006510

Zn-finger, LRP1 type

TUB

IPR000007

Tubby

ZIM

IPR010399

ZIM

DDT

IPR004022

DDT

ZF-HD

IPR006455

Homeobox domain, ZF-HD class

MBF1

IPR001387

Helix-turn-helix type 3

S1Fa-like

IPR006779

DNA binding protein S1FA

CAMTA

IPR005559

CG-1

LFY

IPR002910

Floricaula/leafy protein

Nin-like

IPR003035

Plant regulator RWP-RK

TAZ

IPR000197

Zn-finger, TAZ-type

Potentially novel plant TFs and transcriptional regulators

CCHC (Zn)

112

IPR001878

Zn-finger, CCHC-type

IPR001789, IPR011006

Response regulator receiver

DHHC (Zn)

IPR001594

D or P

Zn-finger, DHHC-type

HTH

FIS

IPR002197

Helix-turn-helix, Fis-type

AraC

IPR000005

Helix-turn-helix, AraC type

BTB/POZ

IPR000210

BTB

TTF-type (Zn)

IPR006580

Zn-finger, TTF-type

IPR001487

Bromodomain

Lambda-DB

IPR010982

Lambda_DNA_bd

TrpR

IPR010921

Trp repressor/replication initiator

TPR

IPR001440

Tetratricopeptide TPR_1

KRAB-box

IPR001909

KRAB box

NRs

IPR008946

LBD

Steroid nuclear receptor, ligand binding

R3H

IPR001374

Single-stranded nucleic acid binding R3H

YEATS

IPR005033

YEATS

U1-type (Zn)

IPR003604

Zn-finger, U1-type

A20-like

IPR002653

Zn-finger, A20-type

Euk_TF

IPR008917

Euk_TF, DNA binding

NGN

IPR006645

NGN

p53-like

IPR008967

p53-like TF, DNA binding

SSB protein

IPR011344

Single-strand binding protein

ssDB TR

IPR009044

Single-strand DNA binding transcriptional regulator

TCoAp15

IPR003173

Transcriptional coactivator p15

BED-type (Zn)

IPR003656

Zn-finger, BED-type predicted

TCoA

IPR009255

Transcriptional coactivation

Tc/PD

IPR001533

Transcriptional coactivator

IMGAG (International Medicago Gene Annotation Group)-proteins were classified as putative TFs if they contained characteristic DNA-binding or other characteristic TF domains and if annotations of matching proteins obtained by BLAST searches were consistent with such a classification. TF families previously identified in plants are presented in the first part of the table while potentially novel plant TF families, which were identified by the presence of domains associated with TFs and other transcriptional regulators outside the plant kingdom, are presented in the latter part of the table. D = DNA binding domain; P = protein-protein interaction domain; NA = nucleic acid (DNA and RNA) binding domain; RD = receiver domain; LBD = ligand binding; transcriptional co-activator. Plant-specific TF families and sub-families are indicated in bold (according to 12)

PCR primer design

To ensure maximum specificity and efficiency during PCR amplification of TF cDNA under a standard set of reaction conditions, a stringent set of criteria was used for primer design. This included predicted melting temperatures (T_m) of 58°C to 61°C, limited self-complementarity and poly-X, and PCR amplicon lengths of 100–150 base pairs (bp). Secondary hits were minimized by aligning primer candidates to all known Medicago sequences via WU-BLAST 28 and eliminating primer pairs with multiple potential hits.

PCR primer testing: gene-specificity and amplification efficiency

PCR primers were tested on Medicago cDNA free of genomic DNA contamination as follows. First, total RNA was extracted from various organs using Trizol reagent (Invitrogen GmbH, Karlsruhe, Germany), which yielded high quality RNA as judged by gel electrophoresis and by Agilent 2100 BioAnalyser using RNA 6000 Nano Chips (Agilient Technologies, Waldbronn, Germany). Typical RNA yields ranged from 0.5–1.0 μg RNA/mg fresh mass for nodules and leaves, respectively. Isolated RNA was treated with DNAse I (Ambion, product number 1907) to remove all contaminating genomic DNA, and this was always confirmed by PCR using primers to non-coding regions of the Ubiquitin gene (TC102473; AC137828-19.4). After inactivation of DNAse I, RNA was reverse transcribed using SuperScript III reverse transcriptase (Invitrogen GmbH, Karlsruhe, Germany) and oligo-dT_12–18to prime the reaction.

Specificity of PCR primers was assessed in three ways: by melting curve analysis of PCR reaction products; by separating the products of all reactions via electrophoresis in 3% agarose gels; and by sequencing a sub-set of PCR reaction products (Figure 1). 94.5% (998/1045) of primer pairs gave unique PCR products of the expected size. Only 3.3% (34/1045) of primer pairs yielded no product and 2.2% (23/1045) gave non-specific products. (see Additional file 1). Sequencing was performed on 178 randomly-chosen PCR products amplified from a 1:1 mixture of leaf and root cDNA. In the vast majority of cases (92.7% or 165/178), the sequence of the PCR product was identical to that of the intended target gene. In 5.1% of cases, the amplicon sequence matched multiple related genes, including the target gene, while in only 2.2% of cases the amplicon sequence did not match the target gene sequence.

Figure 1

Specificity of transcription factor PCR primers

Specificity of transcription factor PCR primers. Specificity was confirmed by dissociation curves with a single peak (A) while double peaks (B) indicated off-target ampification. The derivative of fluorescence intensity is shown on the y-axis. Separation of PCR products on 3% (v/w) agarose gels following electrophoresis (C) confirmed the presence of unique amplicons of the expected size for most reactions. Few reactions yielded no products (indicated by arrow).

Ideally, PCR results in an exact doubling of the amount of dsDNA after each temperature cycle. In practice, however, this is generally not the case because the reactions are less than 100% efficient. Primer sequences can affect PCR efficiency, so we determined the efficiency of each TF primer pair from amplification plots, using LinRegPCR software 31. First, the correlation coefficient derived from linear regression analysis of each amplification plot (e.g. see Figure 2) was used to assess the 'quality' of each reaction, and all reactions with an R²< 0.990 were excluded from further analysis (10.6% of reactions). Next, average PCR efficiencies (E) were computed for each individual primer pair across all analyzed samples. 53.4% (558 TF genes) displayed PCR efficiencies greater than 0.80, while 39.7% (415 TF genes) had efficiencies between 0.51–0.80. Only 2.6% (27 TF genes) had mean E values below 0.4; these all yielded R²< 0.99 in LinRegPCR analysis and mostly represented reactions that lacked detectable fragment amplification (C_T> 40) or that generated unspecific PCR products (Figure 2; see Additional file 1). A similar range of PCR efficiencies were determined for Arabidopsis and rice TF primers previously 1920.

Figure 2

Amplification efficiency of transcription factor-specific primer pairs

Amplification efficiency of transcription factor-specific primer pairs. Typical real-time RT-PCR amplification plots of 384 TF genes (left) and distribution of PCR efficiencies for all 1045 TF primer pairs (right).

Selection of reference genes

Reference genes with stable expression/transcript levels throughout development and in the face of environmental challenge are crucial for the normalization of expression data of other genes. Potentially useful reference genes were chosen based on published data for Medicago (e.g. Msc27 32) and Arabidopsis thaliana 33. The closest Medicago homologues of Arabidopsis genes were identified by BlastN 29. Gene-specific primer pairs for the genes encoding elongation factor 1α (EST317575), glyceralaldehyde-3-phosphate dehydrogenase (MtC00030_GC; CT573421_3.4), β-tubulin (TC106341), Pentatricopeptide repeat protein (TC96273), actin2 (TC107326; AC137836_27.5), Ubiquitin (TC102473; AC137828-19.4), Helicase (CB892427), and the genes PDF2 (TC107161), UPL7 (TC111218), PTB (TC111751), UBC (AW686873), bHLH (CX538576), and UBC9 (TC106312) were designed using the criteria described above (Table 2). The specificity of PCR primers was tested using 18 first-strand cDNAs from six different organs of Medicago (three biological replicates each). All primer pairs produced a single PCR product of the expected size, as shown by gel electrophoresis and unique dissociation curves generated by the PCR machine after 40 cycles (Figure 3). To determine which reference genes were best suited for transcript normalisation, we used the software geNORM 34, which uses pair-wise comparison and geometric averaging across a matrix of biological samples to determine gene expression stability (M; 35). The genes PDF2, PPRep, Ubiquitin, and PTB had the lowest M (greatest transcript stability) and, therefore, were judged to be the best reference genes for this diverse set of developmental samples (Figure 4; Table 2).

Table 2

Medicago reference genes and primers for qRT-PCR

Gene Name

Accession

Number

Forward/Reverse Primer (5'-3')

PCR

Product

Size (bp)

PCR

efficiency

(E)

R ²

β Tubulin

TC106341

TTTGCTCCTCTTACATCCCGTG / GCAGCACACATCATGTTTTTGG

100

1.08

1.00

PPRrep

TC96273

GGAAAACTGGAGGATGCACGTA / ACAAGCCCTCGACACAAAACC

100

0.93

1.00

PDF 2

TC107161

GTGTTTTGCTTCCGCCGTT / CCAAATCTTGCTCCCTCATCTG

100

0.99

1.00

bHLH

CX538576

TAGCGAGTACCATGATGCCAGA / GCGCCTCTTTTGTTTTCAGC

100

0.89

1.00

UBC

AW686873

CTGACAGCCCACTGAATTGTGA / TTTTGGCATTGCTGCAAGC

100

0.96

1.00

PTB

TC111751

CGCCTTGTCAGCATTGATGTC / TGAACCAGTGCCTGGAATCCT

100

0.85

1.00

Ubiquitin

TC102473

AC137828_19.4

GCAGATAGACACGCTGGGA / AACTCTTGGGCAGGCAATAA

100

0.95

1.00

UBC9

TC106312

AC137602_2.4

GGTTGATTGCTCTTCTCTCCCC / AAGTGATTGCTCGTCCAACCC

100

1.13

0.99

Helicase

CB892427

GTACGAGGTCGGTGCTCTTGAA / GCAACCGAAAATTGCACCATAC

100

0.91

1.00

ELF1α

EST317575

GACAAGCGTGTGATCGAGAGATT / TTTCACGCTCAGCCTTAAGCT

100

0.68

0.98

UPL7

TC111218

CCAGTTGTTCTCGTGGTCCATT / CCTCCAATTGTCGCCCAAA

100

0.93

1.00

GAPDH

MtC00030_GC

CT573421_3.4

TGCCTACCGTCGATGTTTCAGT / TTGCCCTCTGATTCCTCCTTG

100

1.04

0.99

Actin2

TC107326

AC137836_27.5

TCAATGTGCCTGCCATGTATGT / ACTCACACCGTCACCAGAATCC

100

1.12

0.99

MSC27

X63872 (M. sativa)

GTTGAAGTAGACATTGGTGCTAACG / AGCTGAGTCATCAACACCCTCAT

100

0.76

0.99

Mean PCR efficiency (E) was determined from three biological replicates of each of six organs, using LinRegPCR 31, which also yielded mean R². N, no corresponding GenBank accession number.

Figure 3

Specificity, efficiency, and reproducibility of PCR primers designed to amplify reference gene transcripts

Specificity, efficiency, and reproducibility of PCR primers designed to amplify reference gene transcripts. Specificity of primers was confirmed by the presence of unique amplicons of the expected size following electrophoresis on 3% (v/w) agarose gels (A) and by dissociation curves with a single peak (B to D). Typical real-time RT-PCR amplification plots of three reference gene transcripts (E to G).

Figure 4

Ranking of 8 reference genes in M. truncatula

Ranking of 8 reference genes in M. truncatula. Transcript levels of all 8 genes were measured by qRT-PCR, using 18 independent cDNA preparations from six different organs with three replicate measurements of each cDNA preparation. A low value for the average expression stability M, as calculated by geNORM software, indicates more stable expression throughout the various organs.

Identification of organ-enhanced TFs of Medicago

To get an overview of TF gene expression in Medicago truncatula and to identify TFs induced in specific organs, we used the real-time RT-PCR platform described above. Transcript profiling was performed on six different organs of Medicago (leaves, stems, flowers, pods, roots, and nodules) with three independent biological replicates for each (see Additional file 2). The fraction of genes for which transcripts were detected within 40 cycles ranged from 77.2% in leaves to 90.8% in pods. Transcripts from nearly all putative TF genes (96.8% or 1011/1045) were detected in at least one organ. Genes were called detected if they were expressed in at least two biological replicates with a C_T< 40. Approximately half of all TF genes exhibited differential expression during plant development, based on significant differences (p ≤ 0.05) in transcript levels between organs. Few TF genes (1.19% or 12/1011) were expressed exclusively in vegetative organs (leaves, stems, roots or nodules), and even fewer (0.5% or 5/1011) were expressed only in reproductive organs (flowers or pods). A relatively small number of TF genes exhibited greater than ten-fold ratios in expression level in one organ compared to any other organ (Table 3). For comparison, we have included gene expression ratios derived from Affymetrix array data from the same RNA samples. While there is reasonable qualitative agreement between gene expression ratios obtained using the two methods, the lack of quantitative agreement is likely due to the limited sensitivity and low signal to noise ratio near the detection limit of Affymetrix arrays 19. The genes listed in Table 3 may control development and/or differentiation in Medicago and are interesting targets for future research.

Additional file 2

Complete list of TF genes, gene families and experimental data. Shown are the Medicago gene Accession Numbers (TIGR) in ascending order, the TC number if available, as well as the transcription factor family and the subfamily names. The next columns show experimental results for the real-time RT-PCR reactions performed on six different organs of Medicago in three indipendent biological replicates, as explained in the manuscript. CT = not normalized CT value, ΔCT = CT value normalized against the geometric mean of 4 house keeping genes; ΔΔCT = power(PCReff;-ΔCT); log2 ΔΔCT = logarithmus of ΔCT.

Click here for file

Table 3

Organ-enhanced TF genes

A Root-enhanced TF genes identified by real-time RT-PCR

Real-time RT-PCR Expression Ratio

Affymetrix Expression Ratio

Accession Number

R/L

R/S

R/F

R/P

R/N

R/L

R/S

R/F

R/P

R/N

Affymetrix Chip ID

AC140721_12.1

121.5

320.1

1623.4

2676.5

64.4

161.2^b

144.5^b

219.2^b

158.1^b

99.9

Mtr.50075.1.S1_s_at

AC135101_25.1

312.2

438.0

291.6

787.0

61.3

AC140721_14.1

1204.9

751.2

956.6

6296.6

32.6

AC140031_3.1

275.4

1157.3

923.4

258916.9^a

31.1

0.4

0.7

0.8

0.9

2.2

Mtr.47227.1.S1_s_at

AC140721_13.1

39.9^a

179.8

614.8

935.1^a

26.5

35.9

48.1^b

41.2

44.2^b

39.5

Mtr.50074.1.S1_at

AC140031_7.1

360.2^a

830.6

1542^a

6868.8

19.7

0.7^b

1^b

1.1^b

1^b

0.9^b

Mtr.47229.1.S1_at

AC146574_6.1

1095.7

11.9 ^a

48.8

51.8

88.8

1.2^b

1.1^b

1.3^b

1.1^b

1.2^b

Mtr.40781.1.S1_s_at

AC125478_13.7

1117.4

7008.0

3264.9^a

7256.6

11.6

211.5^b

248.8^b

296.4^b

228.7^b

45.3

Mtr.15416.1.S1_at

AC125478_7.2

1052.3

503.3

938.4^a

2754.5

11.2

AC122726_21.111

50.0

48.0

20943.9

2658.6

10.4

74.3^b

76.4^b

72.8^b

82.8^b

26.6

Mtr.15568.1.S1_s_at

B Nodule-enhanced TF genes identified by real-time RT-PCR

Real-time RT-PCR Expression Ratio

Affymetrix Expression Ratio

Accession Number

N/L

N/S

N/F

N/P

N/R

N/L

N/S

N/F

N/P

N/R

Affymetrix Chip ID

AC148816_3.2

9966.7

1826.2

6262.9

21709.1

662.7

434.6^b

516.8^b

455^b

521.7^b

551.4^b

Mtr.14503.1.S1_at

AC147774_3.2

55.2 ^a

466.7

726.3

881.2

498.5

17.7^b

16^b

17.8^b

19.1^b

16.1^b

Mtr.19554.1.S1_at

AC138056_33.241

43.9

93.2

18.3

17.3

13.1

0.9^b

0.8^b

1.1^b

Mtr.17993.1.S1_at

AC124214_39.2

121.8

92.1

44.8

64.9

10.5

C Pod-enhanced TF genes identified by real-time RT-PCR

Real-time RT-PCR Expression Ratio

Affymetrix Expression Ratio

Accession Number

P/L

P/S

P/F

P/N

P/R

P/L

P/S

P/F

P/N

P/R

Affymetrix Chip ID

AC143340_4.7

154.1

44.8

173.3

180.3^a

407.3

1.6^b

1.8^b

1.7^b

Mtr.17931.1.S1_at

D Flower-enhanced TF genes identified by real-time RT-PCR

Real-time RT-PCR Expression Ratio

Affymetrix Expression Ratio