Genetic and Bioinformatic Characterization of Puroindoline A and Puroindoline B in Italian Wheat Cultivars | Open Access Journals

ISSN: 2320-0189

Genetic and Bioinformatic Characterization of Puroindoline A and Puroindoline B in Italian Wheat Cultivars

Bruna De Felice*, Francesco Manfellotto and Raffaella D’Alessandro

DISTABIF- Department of Environmental, Biological and Pharmaceutical Sciences and Technologies, University of Campania Luigi Vanvitelli Caserta, Italy

*Corresponding Author:
Bruna De Felice
DISTABIF-Department of Environmental
Biological and Pharmaceutical Sciences and Technologies
University of Campania Luigi Vanvitelli Caserta
Italy
Tel: ++39-823-274543.
E-mail: bruna.defelice@unina2.it

Received date: 19/07/2017; Accepted date: 24/07/2017; Published date: 29/07/2017

Visit for more related articles at Research & Reviews: Journal of Botanical Sciences

Abstract

Wheat kernel texture is a critical property influencing wheat raw material processing and end use products features. Wheat grain texture is a genetically determined and inherited trait. Differences among wheat cultivars reside in variable puroindoline a and b (Pina and Pinb) allelic forms, leading to a continuous range of variable grain hardness. Genetic determination of Pina and Pinb sequences in a large number of international cultivars has led to the assessment of genetic polymorphisms on wheat grain texture. Here we report the identification of Pina-D1 and Pinb-D1 allelic polymorphism in twenty-eight Italian soft wheat cultivars (Triticum aestivum L.) and ten Italian durum wheat cultivars (Triticum turgidum L. ssp. durum). In contrast to previous studies, our results revealed the presence of Pinb-D1b, Pinb-D1d and Pina-D1b (deletion allele) alleles in Italian cultivars. Since only some puroindoline a and b alleles have been investigated for their direct effect on wheat grain texture, we performed a bioinformatic analysis of puroindoline polymorphisms, to predict the possible impact of the biological function of such proteins on wheat. This investigation could provide useful information to extend the current knowledge about genetic determinants of kernel hardness in soft wheat and indications for breeding and food product applications.

Keywords

Puroindoline genes, Wheat, Grain texture, Genetic analysis, Bioinformatics.

Introduction

Cereals are human nutrition main components. Different wheat uses request a classification of wheat properties focusing on kernel texture, as "soft" (or bread) wheat (Triticum aestivum L.) or as "durum" (very hard) wheat (Triticum turgidum L. ssp. durum) [1]. However, soft wheat varieties can be further classified into subclasses due to their different texture hardness, ranging from soft to hard. Wheat kernel texture is a critical feature for wheat classification and food industry destination. The common, hexaploid wheat cultivars (Triticum aestivum L.) are used for breads, cookies, cakes, and pastries, while the very hard tetraploid cultivars, derived from Triticum turgidum var. durum (AABB), are mainly used for Italian-style pastas.

Grain texture is an inherited trait and the principal genes involved in grain texture determination were found to be located in the short arm of chromosome 5D (specifically found in T. aestivum). That locus was named Hardness (Ha) locus and is known to control grain hardness in wheat. The softness phenotype is recognized as the dominant trait [2]. This locus contains the puroindoline a (Pina) and puroindoline b (Pinb) genes (representing the major determinants of wheat kernel texture), which are the subunits of a roughly 13 kDa friabilin, and grain softness protein gene (Gsp-1).

PINA and PINB proteins determine grain texture binding the surface of starch granules in the endosperm cells and forming the previously described protein complex. The soft phenotype is determined by the wild-type alleles, named Pina-D1a and Pinb- D1a, of the Puroindoline a (Pina) and Puroindoline b (Pinb) genes. Pina-D1 and Pinb-D1 genes coding sequences are intronless and 447 bp long sharing 70.2% coding sequence. Such genes encode wheat endosperm-specific lipid binding proteins 148 amino acids long with a cysteine-rich backbone and a unique tryptophan-rich domain, which was considered as being responsible for the strong affinity of the Puroindoline-D1 protein to polar lipids [3].

The hard texture phenotype results from various mutations in either one or both of the puroindoline genes or complete absence of puroindoline coding genes, respectively [4]. Indeed, it has been shown that RNA interference (RNAi)-based silencing of Pina and Pinb genes significantly decreased the puroindoline a and puroindoline b proteins in wheat and essentially increased grain hardness [5]. All wheat cultivars to date that have mutations in Pina or Pinb genes are hard textured, while wheat possessing both the 'soft type' Pina-D1a and Pinb-D1a sequences are soft textured.

The genetic bases of puroindoline absence in durum wheat is primarily determined by the evolutive events that have led to the speciation of durum and soft wheat. Durum wheat (Triticum turgidum) originated about 500,000 years ago from a hybridization event between Aegilops speltoides and Triticum urartu, resulting in a tetraploid species (genome AABB). A subsequent hybridization event (occurred about 8,000 years ago) between Triticum turgidum and Aegilops tauschii resulted in the origin of the hexaploid common wheat (Triticum aestivum, genome AABBDD). Current evidence suggest that a deletion event occurred at Ha locus during hybridization, leading to the complete absence of those genes in durum wheat. Instead, Pina and Pinb genes have been provided to common wheat genome by A. tauschii (D chromosome series) [6]. Conversely, the third gene belonging to the Ha locus, Gsp-1, has been conserved also in durum wheat. The presence of GSP genes and their probable expression in durum wheat suggest a secondary role in determining grain texture [7].

Puroindoline genes arouse scientific interest worldwide for their kernel texture conferring properties. Their investigation in several wheat cultivars has led to the identification of a large number of alleles for both Pina-D1 and Pinb-D1 genes [8]. Following their identification, puroindoline genes have been further investigated for their specific correlation to soft wheat kernel texture [9]. In addition to their texture-related function, puroindolines have also been recognized as a-amylase inhibitors (AAIs) and Seed Storage (SS) proteins, having a possible role in protecting plants against pathogens with their bactericidal and fungicidal activities [10,11].

As the largest producer and consumer of “soft and hard“ wheat in the entire world, Italy is a center of wheat, holding a highly diverse stock of wheat germplasm, proving useful in both applied and basic research efforts to give insight into the biology of wheat plant. Germplasm and the scientific method of breeding provide the foundation for bountiful wheat harvests.

In our research, we aimed to investigate Pina-D1 and Pinb-D1 allele polymorphism in Italian Triticum aestivum and Triticum turgidum ssp. durum cultivars, to evaluate possible genetic differences related to phenotypic variability in kernel texture. For this purpose, we amplified and sequenced Pina-D1 and Pinb-D1 genes from each cultivar and performed bioinformatic analyses to assess possible phenotypic consequences on grain texture.

Methods

Plant Material

Ten Italian Triticum turgidum ssp. durum and 28 Italian Triticum aestivum grain cultivar were provided by (CRA-SCS), Battipaglia (Salerno, Italy) and CRA–SCS (Verona, Italy). For each cultivar, grains were grounded and leaves were used to extract the DNA.

Fresh leaves form wheat cultivars were collected and immediately grinded in liquid nitrogen. Samples were stored at -80°C until DNA extraction was performed.

DNA Extraction from Wheat Leaves

DNA was extracted from 1.6 grams of grinded wheat leaves following the CTAB extraction method [12]. DNA yield and purity was assessed a using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE). Purity of the extracted DNA was based on the 260/280 and on the 260/230 O.D. ratios. DNA integrity was assessed running extracted samples on 1% (w/v) agarose gel stained with ethidium bromide.

PCR Amplification of Puroindoline and Control Genes

PCR amplifications were carried out on each DNA sample. As positive control, Gsp1 gene was amplified (because of its presence also on A- and B- wheat genomes, other than on D-genome). Gsp1 primer sequences were Gsp-1F-5’-CTTCCTCCTAGCTTTCCTTG-3’ and Gsp-1R- TAGTGATGGGGATGTTGCAG-3’ [13]. Primers used for amplification of Pina-D1 and Pinb-D1 were Pina-F-5’- ATGAAGGCCCTCTTCCTCA-3’, Pina-R-5’-TCACCAGTAATAGCCAATAGTG-3’, Pinb-F- 5’-ATGAAGACCTTATTCCTCCTA-3’, Pinb-R- 5’- TCACCAGTAATAGCCACTAGGGAA-3’, respectively [3]. In samples missing Pina-D1 amplification, a second pair of primers was used to assess the presence of Pina-null allele (Pina-D1b). Their sequences were: Pina-D1b-F-5’-AATACCACATGGTTCTAGATACTG-3’ and Pina-D1b-R- GCAATACAAAGGACCTCTAGATT-3’ [14].

PCR reactions were performed in 25 μL total volume containing 25 pmol of each primer, 250 μM of each dNTP, 1x PCR buffer, 1.5 mM of MgCl2, 0,5 units of Taq DNA polymerase (Promega, USA) and 100 ng of DNA template. GSP1, Pina-D1 and Pinb-D1 amplification program comprised an initial denaturation at 95°C for 5 min, followed by 35 cycles of 95°C for 40 s, 55°C for 50s, 72°C for 1min and a final extension at 72°C for 7min. Pina-D1b amplification program comprised an initial denaturation at 95°C for 5min, followed by 35 cycles of 95°C for 40s, 57°C for 50s, 72°C for 1min and 72°C for 7min.

The expected Gsp1 amplicon was 467 base pairs (bp) in length, while Pina-D1 and Pinb-D1 were 447 bp in length and Pina- D1b was 778 bp. PCR amplification was subsequently assessed running the amplified product on 1,2% (w/v) agarose gel stained with ethidium bromide and comparing amplicon size with a DNA size standard for gel electrophoresis (DNA Ladder 100 bp plus, Applichem, Darmstadt, Germany; DNA molecular weight marker XVII 500 bp ladder, Roche Diagnostics, Mannheim, Germany).

Extracted DNA fragments were re-amplified by PCR, cloned using TA Cloning kit (Invitrogen, USA) and sequenced. Each Pina and Pinb PCR product was cut from the gel, purified with the Qiaquick Gel Extraction Kit (Qiagen, Hilden, Germany), cloned into the pMOSBlue vector, pMOSBlue blunt ended cloning kit, (Amersham, Piscataway, NJ, USA), and used to transform the DH5α competent cells (Stratagene, La Jolla, CA). Inserts of the desired size were evaluated using PCR with T7 and M13 primers.

DNA Sequence Analysis

Recombinant clones were sequenced from both strands using the Big Dye Terminator v3.1 Cycle Sequencing Kit via the automatic sequencing system ABI PRISM 3100 automatic DNA sequencer (Applied Biosystems, Foster City, CA).

Restriction Sites Allele Identification

To assess the identity of the specific allele polymorphism a restriction assay was performed on Pinb-D1 amplicons, using SapI restriction enzyme (predicted to cut only the Pinb-D1b allele) (Thermo Scientific, USA). We digested 10 μL of amplicon, 5 units of SapI enzyme, 2 μL of buffer and incubated for 16 hours at 37°C. Inactivation was conducted at 65°C for 2 hours. Gel electrophoresis was performed on a 3% agarose gel stained with ethidium bromide. To detect the two digested fragments, and compared to pBR322 DNA-MspI digest molecular marker (New England Biolabs, Ipswich, MA, USA).

Bioinformatic Analyses

NE cutter V2.0 was used to evaluate restriction enzyme sites maps in each known Pina-D1 and Pinb-D1 alleles with the aim to evaluate differences in the presence/absence or in the position of restriction sites. Such evaluation allowed to perform specific PCR amplicon digestions in order to confirm the presence of allelic variants. The nature of cloned sequences was confirmed by performing computer-based similarity searches with known Pina and Pinb sequences in the NCBI database using BLASTN algorithm [15] to assess the specificity of the amplified product against the entire database.

Direct comparison between wild type puroindoline sequences Pina-D1a and Pinb-D1a and sequenced products was performed using Clustal W program [16] and verifying, for each sequence position, the nucleotide correspondence to the reference allele. The nucleotide sequences were translated using the ExPASy translation tool and translated sequences were submitted to PROVEAN software tool to predict whether amino acid substitutions or indels could have an impact on the biological function of the analysed protein. Cut-off value to predict a protein alteration as deleterious was set at -2.50 [17].

Results

Gsp-1, Pina-D1 and Pinb-D1 Genes Amplification

All analysed wheat DNA samples gave GSP-1 gene amplification (the above mentioned specific 467 bp amplicon), confirming the suitability of extracted DNA for puroindoline genes investigation. Puroindoline gene amplification was confirmed in each soft wheat DNA sample. Figure 1 shows the amplification results obtained for Claudio, Guadalupe and Adelaide cultivars.

botanical-sciences-genes-amplification

Figure 1: Typical Gsp-1, Pina-D1 and Pinb-D1 genes amplification in Triticum aestivum and Triticum turgidum ssp. durum wheat samples. Lane M: 100 bp ladder; Lanes 1, 2 and 3: Claudio durum wheat Gsp-1, Pina and Pinb amplification; Lanes 4, 5 and 6: Guadalupe bread wheat, Gsp-1, Pina and Pinb amplification; Lanes 7, 8 and 9: Adelaide bread wheat, Gsp-1, Pina and Pinb amplifications.

As we did not get any Pina-D1 amplification in some of the analysed cultivars, we performed a PCR assay to specifically evaluate the presence of null Pina-D1 allele (Pina-D1b) (Figure 2).

botanical-sciences-Domain-structure

Figure 2: Domain structure of Pina-D1 null allele variant from Triticum aestivum (Profeta and Tirex cultivar). Deduced amino acid sequence is shown for the 23-nucleotide long Pina coding sequence belonging to Pina-D1b null allele.

The cultivars assessed for the large deletion were all durum wheat cultivars. Moreover, as Guadalupe cultivar was previously reported to lack Pina-D1 amplification, we performed a Pina-D1b specific PCR on that cultivar. Beside Guadalupe, we got Pina-D1b amplicon for Amidon, Profeta, Ciano and Sibilia cultivars [12]. Figure 3 shows Pina-D1b amplification positive results.

botanical-sciences-amplification-Profeta

Figure 3: (A) Pina-D1b (null allele) amplification in Profeta (Triticum aestivum) and six Triticum turgidum ssp. durum wheat samples. Lane 1-10: Profeta and Sibilia (bread wheats), Cirillo, Giemme, Saragolla, Claudio, Core, Creso and Zenit Pina-D1b amplification; Lane M: Marker XVII 500 bp (Roche). (B) Pina-D1b (null allele) amplification in Guadalupe cultivar. Lane 1: Guadalupe Pina-D1b amplification; Lane M: Marker 100 bp.

Pina and Pinb Genes Sequence Analysis

In order to confirm the specificity of Pina-D1, Pinb-D1 and Pina-D1b amplification and the specific allelic form, we excised from agarose gels, cloned and sequenced the amplicons of interest. BLASTN analyses revealed that the obtained amplifications were specific for the expected genes when compared to public sequence databases.

All amplified sequences were aligned with the reference alleles (Pina-D1a and Pinb-D1a) using ClustalW program to evaluate the specific genotype and the translated sequence was obtained using ExPASy translation tool. Among soft wheat cultivars, the most common genotypes resulted Pina-D1a/Pinb-D1b (found in Adelaide, Pandas, Colfiorito and Dorico cultivars) and Pina-D1a/ Pinb-D1d (Aubusson and Bologna), but Pina-D1a-b/PinbD1a (Guadalupe) genotype was also found (we obtained positive results for both Pina-D1a and Pina-D1b amplifications). Besides, five soft wheat cultivars, Amidon, Profeta, Ciano, Guadalupe and Sibilia showed the presence of the Pina-D1 null allele (Pina-D1b/Pinb-D1a genotype).

Table 1. shows the identified Pina-D1/Pinb-D1 genotypes for each T. aestivum and T. turgidum ssp. durum wheat. Pina-D1 sequences were found to be corresponding to the wild type allele (Pina-D1a, Genbank DQ363911) for all remaining bread wheat cultivars, so that the predicted codified amino acidic sequence is identical to the wild type puroindoline a protein. Pinb-D1 amplicons sequencing revealed a higher level of heterogeneity in investigated cultivars, as two different alleles were identified beside the wild-type in Triticum aestivum. Indeed, most analysed cultivars showed the wild type version of Pinb-D1 allele when aligned with the reference sequence (Pinb-D1a, Genbank DQ363913). Their predicted codified protein sequence was corresponding to the wild type version of the protein Pinb.

Wheat cultivar Pina-D1 genotype Pinb-D1 genotype
Triticum turgidum ssp. durum
Claudio none none
Cirillo none none
Cosmodur none none
Core none none
Creso none none
Giemme none none
Grazia none none
Neodur none none
Saragolla none none
Zenit none none
Triticumaestivum
Adelaide Pina-D1a Pinb-D1b
Amidon Pina-D1b(null allele) Pina-D1a
Aubusson Pina-D1a Pinb-D1d
Aurelio Pina-D1a Pinb-D1a
Bolero Pina-D1a Pinb-D1a
Bologna Pina-D1a Pinb-D1d
Centauro Pina-D1a Pinb-D1a
Ciano Pina-D1b(null allele) Pinb-D1a
Colfiorito Pina-D1a Pinb-D1b
Dorico Pina-D1a Pinb-D1b
Francia Pina-D1a Pinb-D1a
Genio Pina-D1a Pinb-D1a
Gladio Pina-D1a Pinb-D1a
Guadalupe Pina-D1a/Pina-D1b Pinb-D1a
Lampo Pina-D1a Pinb-D1a
Leone Pina-D1a Pinb-D1a
Libero Pina-D1a Pinb-D1a
Livio Pina-D1a Pinb-D1a
Mosè Pina-D1a Pinb-D1a
Neviano Pina-D1a Pinb-D1a
Oscar Pina-D1a Pinb-D1a
Pascal Pina-D1a Pinb-D1a
Sagittario Pina-D1a Pinb-D1a
Pandas Pina-D1a Pinb-D1b
Profeta Pina-D1b (null allele) Pinb-D1a
Salgemma Pina-D1a Pinb-D1a
Serena Pina-D1a Pinb-D1a
Sibilia Pina-D1b (null allele) Pinb-D1a

Table 1: Pina-D1/Pinb-D1 genotypes for investigated wheat cultivars.

However, other cultivars showed nucleotide mismatches compared to the wild type allele and, specifically, Colfiorito, Dorico, Adelaide and Pandas showed a sequence corresponding to Pinb-D1b allele (G223A nucleotide change, Genbank DQ363914) and Aubusson and Bologna cultivar showed the presence of Pinb-D1d allele (T217A, sequence not deposited in Genbank,) (Figure 4).

botanical-sciences-allelic-variants

Figure 4: Domain structure of the Pinb-D1b and Pinb-D1d allelic variants from Triticum aestivum. Colfiorito, Dorico, Adelaide and Pandas cultivar showed the presence of Pinb-D1b allele (G223A nucleotide change) while Aubusson and Bologna cultivar Pinb-D1d allele variant (T217A) leading to G75S and W73R aminoacidic changes respectively in the predicted codified protein.

Those nucleotide changes are causative of an aminoacidic change in the predicted codified protein and, in details, Pinb-D1b leads to W73R and Pinb-D1d to G75S amino acidic changes (Figure 5).

botanical-sciences-nucleotide-sequence

Figure 5: Pinb-D1b reference allele nucleotide sequence (Genbank DQ363914) and T. aestivum ssp. durum amplified sequences. Deduced Pinb amino acid sequence is shown. Nucleotide sequences variations underlined and aminoacidic sequences showed.

Pinb-D1b allele amplification was confirmed performing SapI restriction digestions on Pinb-D1 amplicons. Restriction site analyses predicted the presence of a specific site recognized by that enzyme only in Pinb-D1b allele, arising from the polymorphism at nucleotide 223 and resulting in a single cut in position 214 of the amplicon. As expected, only in Colfiorito, Dorico, Adelaide and Pandas Pinb-D1 amplicons were digested and agarose gel electrophoresis showed two fragments of 214 bp and 233 bp (Figure 6).

botanical-sciences-SapI-restriction

Figure 6: Pinb-D1b allele digested by SapI restriction enzyme. (A) Nucleotide position, forward and reverse sequences are reported. SapI cutting site is shown as an interruption in the sequence and nucleotide sequence recognized by SapI is underlined. Pinb-D1b polymorphic site is reported in grey. (B) SapI restriction analysis of Pinb-D1b allele. Lane M: pBR322 DNA-MspI digest molecular marker; Lane 1: Levante Pinb amplicon digestion; Lane 2: Adelaide Pinb amplicon digestion; Lane 3: Pandas Pinb amplicon digestion.

Protein Function Prediction Analysis

Protein function prediction effect of published wheat-related genetic polymorphisms leading to an amino acidic change was assessed using PROVEAN software, after translating nucleotide sequences into their corresponding protein sequences. In this analysis, all known Pina-D1 and Pinb-D1 gene variation determining an amino acidic change in wheat were included.

Bioinformatic prediction analysis has assessed whether PINA and PINB amino acid substitutions or indels had an impact on the biological function of the protein. The prediction was based on the change, caused by a variation, in the similarity of query sequence to closely related sequences collected through BLAST. Prediction results allowed the identification of 5 deleterious Pina-D1 (Pina-D1b, Pina-D1l, Pina-D1m, Pina-D1n, Pina-D1p) and 9 deleterious Pinb-D1 amino acidic alterations (Pinb-D1aa, Pinb-D1ab, Pinb-D1c, Pinb-D1e, Pinb-D1g, Pinb-D1p, Pinb-D1r, Pinb-D1s, Pinb-D1u) (Table 2).

Allele Nucleotide variation Aminoacidic variation Provean Score Protein variation effect Previously reported phenotype*
Wheat-related Pina-D1 alleles
Pina-D1b Deletion from nt 24 to the end Deletion from aa 8 to the end -212.987 Deleterious Hard texture, no PINA protein on starch granules and reduced PINB
Pina-D1k Complete Pina-D1 deletion - n.a. n.a. Hard texture, absence of PINA and PINB proteins
Pina-D1l 265Cdel K89fs -94.340 Deleterious Hard texture, PINA null
Pina-D1m C187T P63S -3.298 Deleterious Hard texture
Pina-D1n G212A W71X -3.909 Deleterious Hard texture, PINA null
Pina-D1p T38A V13E -2.306 Neutral n.a.
  410delC C138fs -11.15 Deleterious  
Pina-D1q C417A, A418C N139K, I140L -2.009 Neutral n.a.
Wheat-related Pinb-D1 alleles
Pinb-D1aa C96A, 213delA K71fs -126.926 Deleterious n.a.
Pinb-D1ab C382T Q128X -2.844 Deleterious n.a.
Pinb-D1b G223A G75S -0.326 Neutral Hard grain texture; low levels of starch surface friabilin
Pinb-D1c T266C L89P -6.496 Deleterious Hard endosperm
Pinb-D1d T217A W73R -1.289 Neutral Medium- hard texture
Pinb-D1e G204A W68X -3.287 Deleterious Hard texture, PINB null
Pinb-D1f G219A W73X -2.226 Neutral Hard texture, PINB null
Pinb-D1g C255A C85X -10.136 Deleterious Hard texture, PINB null
Pinb-D1l A220G K74E -0.693 Neutral n.a.
Pinb-D1p 213delA K71fs -126.926 Deleterious Hard texture, PINB null
Pinb-D1q G218T W73L -2.258 Neutral Hard texture
Pinb-D1r 128insG E43fs -161.281 Deleterious Hard texture, PINB null
Pinb-D1s 128insG, G204A E43fs -161.412 Deleterious n.a.
Pinb-D1t G226C G76R -1.058 Neutral Hard texture
Pinb-D1u 127delG E43fs -166.639 Deleterious Mixed or hard texture
Pinb-D1v G22A A8T -2.243 Neutral Hard texture
C25A L9I -0.879 Neutral  
Pinb-D1w G431T S144I -0.297 Neutral Hard texture

Table 2: Missense Pina-D1/Pinb-D1 wheat alleles and predicted effect on codified protein function (n.a.=not available). Multiple sites mutations in the same polypeptide have been analysed as individual mutations. *in association with the wild type allele of the other puroindoline gene if not otherwise specified. fs=frameshift. Aminoacidic positions are numbered starting from the initial methionine.

As expected, the most deleterious protein alteration resulted Pina-D1b leading to a very large deletion of PINA sequence (only the first seven amino acids are codified by the deleted gene). Conversely, other already known amino acidic changes found in our cultivars (Pinb-D1b and Pinb-D1d), also if leading to an alteration in the protein sequence (G75S and W73R), were predicted to have a neutral phenotypic effect. Restriction enzyme site prediction allowed to identify a unique restriction site in Pinb-D1b variant, i.e., SapI restriction site cutting at nucleotide position 214 of Pinb-D1b amplicon and resulting in a single cut in the amplicon and giving as results 214 bp and 233 bp long fragments.

Discussion

Puroindolines role in determining wheat grain softness has been investigated in recent years. The presence of variations in one or both puroindolines (or the absence of puroindoline proteins) leads to a hard phenotype, ranging from the soft phenotype to the extremely hard phenotype encountered in durum wheat [18]. This model recognizes the softening effect of puroindoline proteins.

Up today, many Pina-D1 and Pinb-D1 alleles have been identified in different geographic bread wheat cultivars from around the world. In wheat 7 different Pina alleles and 17 different Pinb alleles have been described. [19]. The most frequent Pina allele leading to the hard phenotype is a large deletion encompassing most of Pina coding sequence (the deletion starts after the 23th nucleotide from the initial ATG, encompassing 15,380 bp), that is the Pina.D1b allele, while the most frequent encountered Pinb variant is PinB-D1b. In most geographic regions, soft wheat cultivars with the Pinb-D1b allele are commonly found, while Pinb- D1p allele is prevalent in Chinese cultivars [4,20]. Pina-D1 sequence point variations are less frequently encountered, but up to date three SNP alleles leading to an aminoacidic change in the mature protein sequence (Pina-D1m, n and q), a SNP and single nucleotide deletion allele (determining a change in the leader peptide and causing a frameshift in the coding sequence) (Pina- D1p), one single nucleotide deletion allele (Pina-D1l) and one complete gene deletion (Pina-D1k) have been described [21].

Pinb-D1 sequence variations are more numerous and frequently encountered than Pina-D1 alleles. In details, 8 SNPs alleles leading to an amino acidic change in leader peptide or in the mature protein sequence (Pinb-D1b, Pinb-D1c, Pinb-D1d, Pinb-D1l, Pinb-D1q, Pinb-D1t, Pinb-D1v, Pinb-D1w), 4 SNPs alleles leading to a nonsense mutation in the mature protein sequence (Pinb- D1e, Pinb-D1f, Pinb-D1g, Pinb-D1ab), 5 single nucleotide deletions creating a frameshift in the protein sequence (Pinb-D1p, Pinb- D1r, Pinb-D1s, Pinb-D1u, Pinb-D1aa) [22].

In addition to the 17 described sequence variations, complete Pinb-D1 deletion has been described [23]. Even if frequently the hard phenotype encountered in T. aestivum cultivars is determined by alterations in Pina or Pinb sequences, some reports of contemporary deletion of both genes has been described [19]. A comparative analysis of the resulting phenotype expressed when Pina and Pinb are mutated has been provided. The “null allele” Pina-D1b is responsible for a harder phenotype than Pinb-D1b and Pinb-D1p in 2005 and Pina-D1m is reported to induce a harder phenotype than Pina-D1b (null allele) mutation [12,13].

Moreover, cultivars with Pina-null/Pinb-null alleles show the highest hardness index when compared to different combinations of Pina-D1/Pinb-D1 alleles [24]. Therefore, correlation between kernel texture and Pina and Pinb genotype can provide useful information for the improvement of wheat quality, and to extend the current understanding of the molecular and genetic mechanisms determining wheat grain texture. Puroindoline gene expression and protein quantification analyses suggested that the presence of PINA directly affects wheat grain hardness, while PINB influences the same phenotypic feature affecting polypeptide three-dimensional stability and its affinity for polar lipids [25]. According to the critical role of puroindoline a, Pina transcription and protein synthesis have been found to be higher than Pinb in soft wheat [14].

Contrasting data emerged by transcription level analyses of Pina-D1 or Pinb-D1 genes. They have shown no differences in soft and Pina-D1a/Pinb-D1b hard wheat in some studies, while Pina-D1a mRNA expression has been found as reduced in Pinb-D1b hard wheat compared to soft in the research of Capparelli. PINB protein level was drastically reduced in PINA null (Pina-D1b/Pinb-D1a) hard wheat. However, the absence of puroindoline a (determined by both complete absence of puroindoline a gene or Pina-D1b allele) has been previously demonstrated to exert a critical effect in a drastic reduction of puroindoline b expression. Although PINA and PINB are currently supposed to be the principal molecules involved in grain texture determination in wheat, other factors could have a role in that complex trait definition. Indeed, Puroindoline b-2 genes, similarly to Gsp-1, have been identified also in durum wheat, suggesting the presence of a multigene family in the A- and/or B-genome, The current knowledge about puroindoline involvement in wheat texture determination has posed the basis for the comprehension of genetic determinants of this important phenotypic trait, but emerging data about puroindoline gene duplications, puroindoline-related genes identification and their regulated expression control suggest that more complex molecular mechanisms are involved [26].

Italy, among the Mediterranean countries, has the longest tradition in wheat breeding and its germplasm can be considered as one of the richest and most valuable. Since in Italy wheat germplasm has been collected starting from the 1947, our investigation, aimed to identify novel puroindoline gene alleles characterizing Italian wheat cultivars. For the first time, we found that three durum wheat cultivars had null Pina allele (complete absence of amplification for Duilio and Levante). We investigated the presence of null Pina allele in cultivars lacking Pina amplification. The specific study of Pina-D1b allele amplification and the subsequent sequence analysis of the obtained amplicon has allowed the detailed identification of Pina null allele, providing the confirmation that the lack of Pina-D1 gene amplification is due to the presence of that specific allele and no other genomic deletions encompassing Pina-D1 gene. Interestingly, we obtained, for the first time, different results from Corona, which reported lack of amplification of Pina-D1 in Guadalupe cultivar and consequently assigned the Pina-D1b allele. We obtained the amplification of Pina-D1a allele on that cultivar confirmed by sequence analysis. And, when we analysed the specific Pina-D1b amplification, we obtained the expected amplicon for the null allele. We could suppose that the presence of both alleles in Guadalupe cultivar is determined by heterozygosity for the alternative forms of Pina-D1 alleles, which can influence the grain texture [27].

Regarding soft wheat cultivars we identified nucleotide mismatches compared to the wild type allele and, specifically, Colfiorito, Dorico, Adelaide and Pandas showed a sequence corresponding to Pinb-D1b allele leading to W73R an amino acidic change and Aubusson and Bologna cultivar showed the presence of Pinb-D1d allele causative G75S amino acidic changes. Although gene sequence analysis represents the most reliable result for genotyping, the development of simple and fast assays to identify the specific Pina or Pinb allele represent a necessity to extend genetic evaluation to a larger sample and to acquire more information about genetic determinants of grain texture. We evaluated the possibility of Pinb-D1b allele identification through a specific digestion assay (SapI restriction analysis), therefore we performed restriction analysis on Levante, Adelaide and Pandas Pinb-D1b amplicons as they include restriction sites resulting in two fragments visible through gel electrophoresis.

As reported, functional properties of puroindolines reside on their structure. The most critical region for lipid binding is the tryptophan rich loop region (PINA-D1a motif WRWWKWWK; PINB-D1a motif WPTKWWK). To understand the functional effect of Pina and Pinb variants on wheat grain texture, we performed a bioinformatics analysis [28]. Although most of the predictions revealed the deleterious effect of puroindoline sequence alteration in determining the hard texture phenotype, the bioinformatics prediction, based on the comparison of the query sequence with related sequences published in international databases, was not able to identify variations affecting functional aminoacidic regions, such as the tryptophan rich-domain (TRD) of puroindolines., Pinb-D1b, d, are predicted as neutral, although they are located in the region very next to TRD, presumably influencing the functionality of the synthesized polypeptide [29-33].

Conclusion

In conclusion, our study has isolated and characterised in wheat Italian cultivars new allelic polymorphisms. However, the comparison between our bioinformatic results and data emerging from our studies underline the need of functional experiments to validate in silico predictions. Future studies are warranted to clarify the mechanisms underlying the role of Pina/Pinb variations on wheat grain texture.

Acknowledgements

We thank CRA–SCS (Salerno, Italy) and CRA–SCS (Verona, Italy) for providing seeds of wheat species. Author Contributions: BDF conceived and designed the experiments; BDF, MB, FM, and EB performed the experiments. BDF and EB analyzed the data. BDF and EB contributed reagents/materials/analysis tools. BDF wrote the paper.

References