Size of AT(n) Insertions in Promoter Region Modulates Gmhsp17.6-L mRNA Transcript Levels

During earlier experiments, an SSR molecular marker (176 Soy HSP) showing high correlation (70%) with resistance/susceptibility to javanese root-knot nematode Meloidogyne javanica was identified in soybean. After being sequenced, results indicated that the SSR 176 Soy HSP marker was inserted in the promoter region of Gmhsp17.6-L gene. It was also detected in this region that resistant genotypes presented insertions between AT(31) and AT(33) in size and susceptible genotypes, AT(9). Gmhsp17.6-L gene coding region presented a perfect match in amino acid sequence in all soybean genotypes. A ribonuclease protection assay showed that Gmhsp17.6-L gene mRNA transcripts were present in all genotypes. A real-time relative quantification (qPCR) indicated in the resistant individuals higher mRNA transcripts levels, which presented in the sequencing more AT(n) insertions. These results suggest that the number of AT(n) insertions inside this promoter region could modulate up or down gene levels. Those findings can lead to the possibility of manipulating, between some limits, the mRNA transcripts levels using different sizes of AT(n) insertions.


Introduction
Diseases caused by nematodes are among the main factors that contribute to yield losses in soybean, especially in tropical and subtropical regions. It is estimated that 11% of the annual losses in world soybean production result from this parasitism [1]. In Brazil, nematode species of the genus Meloidogyne, especially M. javanica and M. incognita, represent a serious problem in many production areas, such as north of Rio Grande do Sul State, southeast and north of Parana State, south and north of São Paulo State, and south of Minas Gerais State. Also, nematode infections are increasing in soybean expansion areas in central Brazil. Unfortunately, management strategies such as crop rotation are not achieving efficient control of nematode populations. Likewise, a lack of effectively resistant cultivars, with broad adaptation across the country, is contributing to the growing problem [2].
Molecular markers allow us to evaluate resistance genes that show Mendelian inheritance and their use helps to identify genetic polymorphisms, eliminating aspects of phenotypic variation, such as environmental effects. With molecular markers, selections can be made at earlier stages of breeding, reducing costs from nematode population maintenance and inoculation of plants. Microsatellites, also called SSR ("simple sequence repeats"), correspond to DNA sequences with a few base pairs (2)(3)(4)(5)(6) in length, repeated in tandem, such as (AT) n , (ATT) n , (GCC) n , and so forth. The variation in the number of these repeated elements generates variability that can be used to identify polymorphisms in different genotypes [3].
Silva and coworkers [4], in previous genetic studies involving M. javanica resistance, found that the amplification or nonamplification of the microsatellite marker 176 Soy HSP had a high significant correlation with the number of galls observed on the roots of soybean plants, obtained 2 Journal of Biomedicine and Biotechnology from resulting crosses between resistant and susceptible genotypes. This marker belongs to the F linkage group and is located between the RFLP markers A186D and A757V. In another study, Fuganti and collaborators [5] confirmed that in this linkage group, in which other quantitative trait loci (QTLs) conditioning resistance to nematodes have already been reported [4,6], the SSR loci 176 Soy HSP and Satt 114 showed significant correlation with number of galls observed on soybean roots inoculated with the pathogen. 176 Soy HSP and Satt 114 markers explained 46% and 43%, respectively, of phenotypic variation in number of galls per root and 70% of variation on average score on nematode infection.
Based on these previous results obtained by Fuganti and collaborators [5], the 176 Soy HSP microsatellite molecular marker was selected to be further studied in soybean plants, resistant and susceptible to M. javanica when inoculated and noninoculated with the pathogen. The resulting sequence for this molecular marker showed high similarity with a promoter region for a gene encoding a small heat-shockprotein (sHSP) found in soybean, Gmhsp17.6-L (Genbank accession no. M11317).
These proteins, although identified initially in response to heat stress, are now known to be induced by many types of environmental stress, including drought, freezing, and high salinity [7]. As molecular chaperones, they are involved in folding and refolding of proteins during their synthesis and transport, as well as the association of polypeptides with each other and other macromolecules to form oligomeric complexes. They comprise a diverse group of proteins mediating the correct assembly of polypeptides but are not themselves part of the functional assembled structures [8].
The genes encoding HSPs (hsps) are highly conserved and many of these genes and their products can be assigned to families on the basis of sequence homology and typical molecular weight: hsp110, hsp100, hsp90, hsp70, hsp60, hsp40, hsp10, and small/low molecular weight (smHSPs/LMW) hsp families [9]. In higher plants, six nuclear gene families encode an LMW protein which corresponds to their location within distinct cellular compartments, including the cytoplasm, plastids, rough ER, and mitochondria [10]. Additionally, specific smHSPs are expressed during various phases of plant development [11].
In soybean, the genes encoding the small molecular weight HSPs are the predominant class of HSPs synthesized. These proteins are regulated largely at the transcriptional level. Promoter elements, designated as heat-shock elements (HSEs), are located upstream of a TATA box and are responsible for the dramatic thermal induction of HS genes. And although the predominant mechanism of activation of these genes is mediated by binding of the HS transcription factors (HSFs) to the HSEs [12], the induction of hs promoters also involves independent cis elements, as AT-rich sequences, which present potential to influence transcriptional activity.
Some plant proteins bind these AT elements and the protein/DNA complexes when fractioned by electrophoresis are grouped as high-mobility complexes (HMCs) and low mobility complexes (LMCs) [13]. HMC proteins anchored to AT-rich sequences in the DNA facilitate macromolecular assembly by charge blocking [13]. This affinity for DNA targets the microenvironment in which HMCs exert their influence, controlling gene expression by facilitating assembly of the functional initiation complex for RNA polymerase(s) and excluding histones in order to facilitate HSF binding upon induction [13].
Thus, considering the biological function of these ATrich elements in heat-shock promoters and the significant differences found in this region, between the genotypes tested, our objective was to analyze Gmhsp17.6-L (Genbank accession no. M11317) mRNA expression level, using the ribonuclease protection assay (RPA) and real-time quantitative PCR (RT-qPCR), in an attempt to elucidate how genetic differences the samples may influence in the resistance and/or susceptibility responses in plants to the javanese rootknot nematode, M. javanica.

Plant Material.
Soybean genotypes BRS133 (susceptible to M. javanica) and PI595099 (resistant) were used as parents and from a population derived from this cross, lines JF7002, JF7027, and JF7056 were selected from resistant population and lines 256-S, 259-S, and 266-S were chosen from susceptible one. These F 10 lines were selected based in two evaluations, where soybean plants were infected with 3.000 eggs of M. javanica and nematodes were quantified 30 and 72 days later. The numbers of galls per plant were determined consisting in one evaluation. The other was performed scoring nematode infection from 0 (total absence of galls) to 10 (abundant) [14].
Genomic DNA was extracted from leaves of samples, following the protocol proposed by Keim et al. [15], quantified and checked for integrity.

SSR Molecular Marker Amplification, Cloning and
Sequencing Fragment. Fuganti and coworkers [5], searching for molecular markers to assist genotype selection, identified the SSR marker 176 Soy HSP as highly linked to soybeanplant response to javanese root-knot nematode, M. javanica. Since this SSR marker explained 46% and 70%, respectively, of phenotypic variation in number of galls per root and average grade on pathogen infection, amplification product was cloned according to Ausubel et al. [16] and sequenced following the recommendations of Sambrook et al. [17].
The sequence obtained from the SSR marker 176 Soy HSP was analyzed for similarity with known sequences in the NCBI Blast [18]. It showed high similarity with a sequence region inside the small heat-shock protein (sHSP) promoter gene, found in soybean, Gmhsp17.6-L (Genbank accession no. M11317). Using the Gmhsp17.6-L gene sequence available in Genbank, new primers (176SoyHSP F 5 TTT T-TG TTT AAG TTA CTG TAC TGT3 and 176SoyHSP R 5 GCT AGT CTT CTA CAA CCT TCT A3 ) were designed that allowed the amplification of PCR fragments in both genotypes, resistant and susceptible. shock proteins have been known to be involved in responses to stress, including biotic stresses such as nematode infection, it seemed reasonable to assume that genetic differences in the promoter of Gmhsp17.6-L may have an impact on nematode resistance associated with this SSR. As a result, a set of primers (pSoyHSP F 5 GGG CTG CAG GAA TTC TG-A AAT TGG GTC TTT TTG3 and SoyHSPCL R 5 CCC-CCC GGG TTA ACC AGA GAT TTC TAT AGC CT3 ), with sites for restriction enzymes to facilitate the cloning process, was designed to amplify, clone and sequence the entire Gmhsp17.6-L gene.
DNA from parental genotypes PI59099 and BRS133, and from all individuals from both populations (256-S, 259-S, 266-S, JF7002, JF7027 and JF7056), was used in PCR reactions, which were carried out in a Perkin Elmer 9600 thermocycler using the following reagent concentrations After electrophoresis, the amplification product was checked using agarose gel (0.7%) electrophoresis prepared with TBE (Tris base, boric acid, EDTA) buffer 1x and stained with ethydium bromide. All resulting fragments were cloned according to Ausubel et al. [16], sequenced following the recommendations of Sambrook et al. [17], and analyzed for similarity with known sequences in the NCBI Blast [18].

Ribonuclease Protection Assay (RPA).
Due to the significant sequence differences detected among resistant and susceptible soybean genotypes, an RPA was performed with the objective of detecting whether these differences caused inactivation of Gmhsp17.6-L expression in some genotypes. Thus, total RNA was extracted using Trizol reagent (Life Technologies) from resistant PI595099 and susceptible BRS133 parental genotypes. Half of the samples were treated with M. javanica eggs (designated "inoculated") and the other half was not (designated "noninoculated").
As no difference was observed in the CDS region of the resistant and susceptible genotypes (data not shown), the Gmhsp17.6-L coding sequence was used as a probe template, flanked in one strand by the T7 bacteriophage promoter and in the other, by the SP6 promoter, as described in the pGEM-T Easy Vector System I (Promega Corp.).
MAXIscript In Vitro Transcription Kit (Catalog no. AM1308-AM1326, Ambion, Inc.) was used, as described in the instruction manual, to produce complete antisense transcript runoffs (RPA2 F 5 GAC ATC ATC AAA CAA -GAG AA3 and RPA2 R 5 TCT CTC CGC TAA TCT GA-A3 ) from Gmhsp17.6-L CDS inserted in the pGEM-T Easy Vector. After labeling with 32 P-CTP (cytosine triphosphate), only full-length probes were purified from a denaturing gel (5% polyacrylamide/8 M urea/1x TBE) according to the manufacturer's protocol. A positive control, provided with the kit, was used, consisting of DNA from a 250 bp Kpn l-Xba l fragment of the mouse β-TRI-actin gene, subcloned from pAL41 (Alonso et al., 1986).
Two digestion control tubes were prepared, with yeast RNA. After hybridization, ribonuclease was added to the mixture in all experimental samples and to one tube of yeast RNA control tube. To the remaining yeast-RNA control tube, only RNase digestion buffer (without RNase) was added. Inactivation/precipitation process was then carried out and protected fragments were separated in a denaturing gel (5% polyacrylamide/8 M urea/1x TBE) and after electrophoresis were detected by exposure to X-ray film.

Quantitative Real-Time PCR (RT-qPCR).
Using the Gmhsp17.6-L gene sequence available from Genbank, specific primers (SoyHSP PSC F 5 GCTGTGTGTCATTGT-CATCGAA3 and SoyHSP PSC R 5 CAC GGT CTA TTT-CTT GCC TAC ATC3 ) for RT-qPCR were designed with Primers Express Software v2.0 package (Applied Biosystems). These primers amplified an 80 bp fragment, from the gene sequence, after the stop codon.
The biological material used comprised the parental genotypes PI595099 and BRS133, three susceptible lines (256-S, 259-S, and 266-S), and three resistant (JF7002, JF7027, and JF7056) lines from the population. After 8 days in a growth chamber, seedlings were transferred to a greenhouse and inoculated with approximately 660 J 2 juvenile Meloidogyne javanica nematodes per plant. After 1, 3, and 6 days of infection, roots from inoculated and non-inoculated treatments were collected, frozen in liquid nitrogen, and stored at −80 • C.
Using Trizol reagent (Life Technologies), total RNA from samples was extracted and quantified. RT-qPCR was conducted by reverse-transcribing 1.5 μg of total RNA with random primers and M-MLV RT (Invitrogen) according to the manufacturer's instructions. To optimize the cDNA concentration, an efficiency curve was constructed, with different template concentrations.
RT-qPCR reactions were performed with an ABI PRISM 7300 Real Time PCR system (Applied Biosystems) in 96-well reaction plates. All reactions were run in technical triplicates, consisting of 8.0 μL of MilliQ water, 0.5 μL ROX active reference dye, 12.5 μL of Platinum SyBR q-PCR SuperMix-UDG (Invitrogen), 1 μL of forward and reverse primers, and 2 μL of cDNA, which consisted of bulked samples from all three collection days, from inoculated and non-inoculated plants. Ribosomal 18S rRNA gene (Genbank accession no. XO2623.1) was used as an endogenous control [19]. The cycling parameters used were as follows: 50 • C/2 minutes, and 95 • C/2 minutes, and forty-five cycles of 95 • C/15 seconds, 62 • C/30 seconds, and 72 • C/30 seconds. Following the final amplification cycle, a melting curve was acquired by heating to 95 • C/15 seconds, cooling to 60 • C/30 seconds, and slowly heating to 95 • C/15 seconds.
All analyses were done using the SDS software package (Applied Biosystems) using the ΔΔCt method. For parental RT-qPCR analysis (Figure 7), genotype PI595099 non-inoculated was chosen as a calibrator (value 1X). For population analyses (Figure 8), susceptible line 256-S noninoculated was selected as a calibrator (value 1X), as this sample showed the smallest AT (n) insertions inside the Gmhsp 17.6-L promoter region. The calibrator sample serves as the basis for comparative gene expression among all the samples.

SSR Marker 176 Soy HSP PCR Amplification and Sequenc-
ing of GmHSP17.6-L Gene. The QTL analysis revealed the presence of at least one gene located next to the 176 Soy HSP marker, with an LOD (likelihood of odds) score of 27.5 ( Figure 1). The LOD value measures the probability that data were generated due to the presence of a quantitative trait loci (QTL).
The sequenced fragment obtained from the 176 Soy HSP microsatellite marker, amplification of which occurred only in susceptible individuals (Figure 2), consisted of 95 bp, including the sequences of the SSR primers. A search for similarity with other known sequences revealed that this fragment showed 100% homology to the promoter region of a small heat-shock protein found in soybean and deposited in Genbank as Gmhsp17.6-L (accession no. M11317) (Figure 3).
Using the available gene sequence, new primers were designed with the objective of generating bands and also sequencing the gene sequence in all resistant and susceptible samples ( Figure 4). All resulting fragments were cloned and sequenced; all tested individuals aligned and showed high similarity with the Gmhsp17.6-L gene sequence, including promoter ( Figure 5) and CDS regions (data not shown). Amino acid sequence alignments were performed and no mismatches were found among the parental genotypes. Individuals of both populations when compared to and with the Gmhsp17.6-L sequence available from Genbank (data not shown) also did not show amino acid sequence difference. However, in the promoter region, differences between genotypes were found, especially concerning AT (n) repetitions. Thus, sequences obtained from resistant samples presented higher numbers of AT (n) repetitions, including parental (PI595099-AT (32) ) and population individuals (JF7002-AT (33) , JF7027-AT (32) , and JF7056-AT (31) ) when compared with sequences from susceptible individuals, parental (BRS133-AT (09) ) and population individuals (256-S and 266-S-AT (09) ) , and also to the sequence from Gmhsp17.6-L gene (AT (15) ), deposited in Genbank. Although, initially evaluated in the greenhouse as susceptible, individual 259-S presented slower mobility PCR fragments consistent with resistant lines and, consequently, showed more AT (n) repetitions inside Gmhsp17.6-L promoter region, similar to the resistant samples.  (BRS133) parents, in inoculated and non-inoculated treatments, expressed the small heat-shock transcript, encoded by Gmhsp17.6-L gene ( Figure 6). In Lines 01 and 02 susceptible parental BRS133 in inoculated and non-inoculated treatments, respectively, is shown while resistant parental PI595099, in inoculated and non-inoculated treatments, is presented in Lines 03 and 04, respectively.

Discussion
The presence of a gene located next to the 176 Soy HSP marker, in soybean's F linkage group, indicates that this region can play an important role in nematode resistance, as already described in the literature; various authors have shown significant correlations between molecular markers and phenotypic responses related to nematode infection, such as infection score and number of galls per plant [4][5][6].
After being sequenced, this SSR marker showed high homology with the Gmhsp17.6-L promoter region. Although this molecular marker at first was amplified only in susceptible samples, a set of different primers was able to amplify it in resistant samples as well; why the amplification does not occur in resistant plants, even though homology for the SSR primers exists, is a matter for further discussion and study. A possible explanation for the nonamplification of 176 Soy HSP marker is that the superior number of AT repetitions found, in the promoter region of resistant plants, might form a hairpin structure, for example, making difficult the amplification process by the Taq DNA polymerase, consequently inhibiting the formation of amplicons.
The gene alignment ( Figure 5) from fragments from susceptible and resistant plants showed that resistant individuals have a larger microsatellite insertion inside the Gmhsp17.6-L promoter. Thus, resistant individuals presented JF7002-AT (33) , JF7027-AT (32) , and JF7056-AT (31) , whereas susceptible individuals presented AT (09) . The whole coding region alignment also proved that the main difference among all genotypes tested occurs only in the promoter region as no differences were observed in nucleotide sequence or in the amino acid sequence alignment (data not shown).
Based on these preliminary results, it was first postulated that the microsatellite insertion present in the Gmhsp17.6-L promoter region could be switching off the Gmhsp17.6-L gene in susceptible plants since the 176 Soy HSP marker was amplified only in these individuals. However, our findings reveal that both resistant and susceptible plants have the microsatellite insertion. Therefore, the number of AT repetitions inserted in the promoter region may be influencing soybean response to the nematode by controlling SOYHSP176 11317-Gmhsp176-L gene sequence, available at the Genbank. PCR amplification was conducted with primers, which delimitated promoter region. In light gray, AT (n) insertion, in dark gray some heat shock elements, with consensus core sequences 5 AGAAnnTTCT3 , 5 cTTCtaGAAgcTTCtaGAAg3 , and 5 CTnGAAnnTTCnAG3 and underlined probable TATA box.
gene expression, inactivating, activating, or even altering the level of the protein controlled by this promoter. The RPA was performed to test the hypothesis that the length of the AT insertion in the promoter region altered Gmhsp17.6-L mRNA transcripts expression. Results demonstrated that all genotypes in all treatments expressed the heat-shock protein mRNA transcripts ( Figure 6). This result actually should have been expected as the literature indicates that plant heat-shock proteins function as molecular chaperones, protecting cells from protein denaturation resulting from environmental stress (Zhu et al. [20]). Thus, our next hypothesis was that the insertion inside the promoter could be affecting gene expression. To examine this possibility, a relative quantification real-time PCR was performed, as no measurable quantitative differences were detected among samples using the RPA technique.
Results from RT-qPCR showed that when comparing the Gmhsp17.6-L expression pattern between resistant and susceptible samples-parental genotypes and populationshigher levels of Gmhsp17.6-L heat-shock protein transcripts appeared in resistant samples inoculated with the pathogen (Figures 7 and 8). These same individuals presented longer AT (n) insertions inside the promoter, which could be targeted to differential mRNA transcript expression.
Also, it was observed that when susceptible-population sample 256-S, which presented the smallest AT (n) insertion  on the promoter region, was used as calibrator (value 1X), all resistant individuals presented higher gene-transcript levels, even when non-inoculated, which implies an evolutionary mechanism in nematode resistance; even in the absence of the pathogen, resistant genotypes express higher level of Gmhsp17.6-L mRNA transcripts. Similarly, the highest level of expression of gene Gmhsp17.6-L was detected in resistant individual JF7002, which also showed the longest AT (33) repetition, although, in general, all resistant samples showed high Gmhsp17.6-L transcript levels. However, analysis of the heterozygous susceptible individual 259-S, which had an AT (n) insertion even longer than those presented by resistant individuals, showed mRNA transcripts level on RT-qPCR as low as susceptible samples. These data suggest a length limit for AT (n) insertion, apart from which there is no more Gmhsp17.6-L gene induction   or activation. However, new studies on this heterozygous individual need to be done in an attempting to elucidate how AT (n) acts in this specific case. Expression of HSPs is controlled by HSFs (heatshock factors), which is the central mechanism controlling response to heat stress. HSFs exist as inactive proteins mainly in the cytoplasm. Stress causes activation with oligomerization and, eventually, recompartmentation to the nucleus, where they bind to target sequences (HSEs) inducing the expression of genes responsible for heat stress response [12]. In addition to HSEs (heat-shock elements), a number of sequence motifs have quantitative effects on the expression of certain heat-shock genes. In plants, there is evidence for involvement of CCAAT-box elements, GAGA-sequences, ATrich sequences, and scaffold-attachment regions [21][22][23].
Specifically, independent cis elements, AT-rich sequences, potentially influence transcriptional activity. In general, plant proteins that bind to these elements are grouped in two categories based on relative mobility of protein/DNA complexes when fractionated by electrophoresis: high-mobility complexes (HMCs) and low-mobility complexes (LMCs) [13].
In particular, HMC proteins anchored to AT-rich sequences in the DNA facilitate macromolecular assembly by charge blocking. A conserved amino acid is involved in the binding of this class of proteins to double-stranded AT-rich elements. The putative DNA-binding domain consists of a glycine-arginine-proline sequence flanked in the N-terminal direction by several basic amino acids [24]. Based on this proposed model, HMCs binding to AT-element sequences of genes may facilitate the integration of transcription factors into the preinitiation complex of the TATA proximal domain. These proteins also control gene expression by facilitating assembly of the functional initiation complex for RNA polymerase(s) and excluding histones in order to facilitate HSFs binding upon induction [13].
In this context, and when resistant individuals inoculated with the pathogen showed longer AT (n) insertions inside the promoter region, when compared with non-inoculated individuals, we hypothesized that the higher levels of Gmhsp17.6-L gene mRNA transcripts presented by these samples may result from more-effective integration of proteins with transcriptions factors forming preinitiation complexes. We hypothesized that when a longer binding sequence AT (n) is presented, combined with histone exclusion and binding of more induced HSFs, higher levels of gene transcription result. Based on these data, new studies are being planned in an attempting to prove, in vivo, that the higher number of AT (n) repetition inside the promoter region leads to higher gene transcription and protein expression.