Chapter 2: Isolation, localization and Sequence Analysis of an Alfalfa HSF

2.1 Introduction
In Ritossa’s pioneering 1962 paper, which is the first known account of the heat shock response, he described the nature of the soon to be named heat shock genes; “…our experiments could be interpreted in the sense that function of genes is reversible and dependent on environmental conditions.” Ritossa was alluding to characteristic induction and attenuation of genes involved in the heat shock response that he observed in the salivary cells of Drosophila. He and his colleagues, however, were stumped, despite further experiments, as to the mechanism cells use to activate these genes. It wasn’t until yeast biologists became involved in the story that sufficient tools were available to isolate and identify the control mechanism governing these heat induced genes.

By 1988 the control mechanism was identified and the first known heat shock transcription factor (HSF) was isolated from the yeast (Sacromycies cerveciae) (Sorger and Pelham, 1988). Two years later the first animal HSF was cloned from Drosophila (Clos et al., 1990) and the first plant HSFs were cloned from tomato (Scharf et al., 1990). Using homologous probes new HSFs were cloned from numerous species and the list continues to grow.

The HSF exist as a family of transcription factors that activate and attenuate the expression of heat shock genes. Plant HSFs being more numerous and complex than animal HSFs are classified using a nomenclature different from animals, consisting of three main classes (A, B & C) and several subclasses (Table 1.2). The classification system uses amino acid sequence analysis of several conserved regions to fit members into one of the three main classes (see Chapter 1, 1.3). Based on the Arabidopsis thaliana genome, there are fifteen known members from class A, five members from class B and one member from class C (Table 1.2). Within class A, HSFs are further organized into 9 subclasses, some of which are organized even further and are designated with lower case letters (Table 2) (see Nover et al., 2001, Table 1.2, for a complete list of all known plant HSFs).

The Arabidopsis subclass A4 has two distinct members (designated A4a and A4c) based on subtle differences in sequence structure. These two HSFs are nearly identical in the DNA binding domain and oligomerization domain, but exhibit minor differences in the C-terminal activation domain (CTAD). Both have two AHA motifs but their location in the CTAD is different. In Arabidopsis, AtHSFA4a is a confirmed transcriptional activator but is marginally responsive to heat stress (Czarnecka-Verner, personal communication). AtHSFA4c has yet to be confirmed as a transcriptional activator but northern analysis has revealed it to be responsive to heat stress (Nover et al., 2001).

Approaches to cloning new HSFs have followed one of two methods; either hybridization screening or PCR screening using E.coli or lambda phage genetic libraries (cDNA or gDNA). Depending on the sequence divergence between species, the PCR approach tends to have fewer successes. However, the time saving advantages make the PCR approach a more desirable first attempt over the hybridization screening approach.

To study HSF activity in alfalfa, I required a native HSF. I specifically wanted to study the activator type or class A HSFs; however, due to the sequences available at the time (January 1999), the only class A plant HSF accessible to me was the Arabidopsis thaliana HSF AtHSFA4 clone (accession number U68561). I used this cDNA sequence to generate a hybridization probe to probe to obtain an alfalfa homologue. Described in this chapter is the isolation of a class A alfalfa HSF, its comparison to known class A4 HSFs and its localization in the alfalfa genome.

2.2 Materials & Methods

2.2.1 32P Probe Preparation
PCR product from the open reading frame (ORF) of the AtHSFA4 cDNA (formerly known as AtHSF21; accession number U68561; provided by Dr. Eva Czarnecka-Verner, University of Florida) was prepared and gel purified. 32P was incorporated into the template AtHSFA4 DNA using the PrimIt II Labeling Kit™ (Stratagene), which employed the use of the Klenow fragment of DNA polymerase. The labeled probe was then purified from the unincorporated radio nucleotides using NucTrap™ Probe Purification Columns™ (Stratagene, cat# 400701). A 1:100 dilution of the purified probe was prepared in 2x SSC to calculate the percent incorporation of 32P and the specific activity of the probe. A 1 ul aliquot of diluted probe was spotted on DE-81 Whatman filters in triplicate and allowed to air dry for 10 min. The filters were washed 3x for 5 min each in phosphate buffer (0.5 M NaH2PO4, 0.5 M NaHPO4), briefly rinsed in distilled H2O and dried under a heat lamp for 10 min. The filters were then assessed in a scintillation counter and the AtHSFA4 probe had a specific activity of 2.3x109 cpm/ul.

2.2.2 Lambda Phage cDNA Library Construction & Preparation
A cDNA library, provided by Dr. Jeff Volenec and Dr. Susan Cunningham, Purdue University was generated from cold acclimated alfalfa, cv Norseman, field grown bud tissue harvested in September, November and December, 1997. Tissue was frozen and total mRNA was isolated, reverse transcribed and cloned into a Lambda ZAP™ phage library as according to the protocol supplied by the manufacture (Stratagene).

The phage library was grown using XL1-Blue host E. coli in LB medium containing 2% maltose (w/v) and 10 mM MgSO4. Serial dilutions of the library, in SM medium (Appendix B), ranging from 10-3 to 10-8 were prepared to determine the titer of the library. In a culture tube, 0.1 ml of phage dilution was combined with 0.3 ml of XL1-Blue host E.coli cells (OD 1.0) and incubated at room temperature for 20 min to allow the phage particles to adhere to the bacteria. The phage/bacteria inoculate was then incubated at 37°C for 15 min to allow the phage to infect the bacteria. Next, 2.5 ml of 48°C, molten agarose was added to the inoculate and poured onto pre-warmed (40 °C) 80 mm LB plates. The plates were allowed to cool and were incubated overnight at 37°C. The titer of the library was found to be 1.32x1010 plaque forming units (pfu) per ml.

2.2.3 Lambda Phage cDNA Library Screening Transferring the Library
A phage dilution was prepared to obtain 50 000 pfu (plaque forming units). In order to accommodate the entire library, 1x106 plaques were required; 20 plates containing 50 000 pfu/plate were therefore prepared. In addition, to allow for a greater plaque dispersion 1500 mm plates were used instead of the standard 80 mm plates.
A 0.3 ml aliquot of phage dilution was combined with 0.9 ml of XL1-Blue host E.coli cells (OD 1.0) and incubated at room temperature for 20 min then incubated at 37°C for 15 min. Next, 8 ml of 48°C, molten agarose was combined with the inoculate and poured onto pre-warmed (40°C) 1500 mm LB plates allowed to cool and incubated for 10 hrs at 37°C. Once the plaques reached an optimal size, approximately 2-3 mm/plaque, the plates were transferred to 4°C to halt plaque growth.

Phage plaques were transferred to nylon membranes (Millipore Immobilon NY+ transfer membranes cat# INYC 13750) by a plaque lifting procedure. All plaque lifts were performed in duplicate to improve the accuracy of the hybridization analysis. The plates were pre-chilled to 4°C prior to the plaque lifts. For each plate, two membrane lifts were prepared individually. The first membrane was overlaid on the plate for 2 min and removed followed by the second membrane which was overlaid for 4 min. While each membrane was on the plate, membrane orientation was marked using an 18 gauge needle and liquid ink which made identical imprints on the plate and both membranes. The membranes were placed plaque side up on filter paper (Watman) saturated with in denaturation solution (1.5 M NaCl, 0.5 M NaOH) in a glass dish for 2 min. The membranes were then floated plaque side up in neutralization solution (1.5 M NaCl, 0.5 M Tris-HCl pH 8.0) for 5 min and gently rinsed (<30 seconds) in wash solution (0.2 M Tris-HCl pH 7.5, 2x SSC Buffer). Excess wash solution was blotted off using filter paper and the phage DNA was fixed to the membrane by UV cross-linking at 120 000 uJ for 30 sec. The membranes were then allowed to air dry for 10 min, wrapped in plastic wrap, and stored at 4°C. Plaque Hybridization
The probe specificity and activity were confirmed by a dot blot hybridization using the AtHSFA4 plasmid. Plasmid dilutions ranging from 25 ng to 25 pg were prepared and fixed to nylon membranes as described above. All hybridizations were performed as described below.
Each membrane was pre-rinsed in distilled H2O then in pre-hybridization solution (6x SSC, 5x Denhardts solution, 0.5x SDS, 0.05 % sodium pyrophosphate, 100 ug/ml herring sperm DNA) and placed in heat-sealable plastic bags in stacks of seven. Pre-hybridization solution, 20 ml, was added to each bag, sealed and incubated at 37°C for 3 hrs. The pre-hybridization solution was discarded and replaced with 30 ml of hybridization solution (6x SSC, 1x Denhardts solution, 0.5x SDS, 0.05 % sodium pyrophosphate, 100 ug/ml herring sperm DNA, 2x106 cpm of denatured DNA probe) resealed and incubated at 50°C for 48 hrs. This low incubation temperature was necessary due to the low sequence homology of the probe.
Post hybridization, the hybridization solution was poured off and retained at 4°C for later use. The membranes were separated and washed 3x for 5 min each in rinse solution (6x SSC, 0.5% sodium pyrophosphate). The membranes were mounted on filter paper plaque side up, overlaid with tracing paper and placed into exposure cassettes with X-Ray film and intensifying screens. The cassettes were stored at -80°C, allowed to expose for 3 days and then developed. Positive Plaque Identification
Exposed membranes produced autoradiograms exhibiting a scattering of dots, most of which were background and random hybridization. To counter this problem, autoradiograms of replicate membranes were overlayed on a light box and orientation markers were lined up. Positive plaques were identified as those dots that appeared in both replicates in the same orientation. On a light box the exposed film with a positive dot was oriented with its corresponding 1500 mm plate. The plaque was cored out with a glass pipette and transferred to 0.5 ml of SM medium and 20 ul of chloroform. The cored plaque was then incubated at 37°C for 2 hrs to release the phage particles from the agarose. The phage was then re-plated using 80 mm plates and re-hybridized as described above with the exception of using 80 mm membranes. This re-plating and re-hybridization was performed two more times to insure the plaque was derived from a single phage. By the third hybridization nearly all the plaques (~95%) were positive.

The purified and isolated plaque was cored out with a glass pipette, added to 0.5 ml of SM buffer and 20 ul of chloroform and incubated at 37°C for 2 hrs, again to release the phage particles. The positive purified phage was combined with helper phage (provided with the aforementioned library kit), which was used to inoculate a culture of “Solar cell” bacteria (OD 1.0). The solar cells in the presence of the Helper Phage excised and ligated the phagmid plasmid from the Lambda Zap phage vector. Inside the solar cells bacteriophage are inhibited from producing more Phage which ensured the Lambda Zap vector remained as a plasmid. As a result, solar cells are created that contained just the plasmid form of the cDNA insert. Solar cell containing plasmid were selected in the presence of ampicillin. The plasmid was then extracted and sequenced using the T3 and T7 flanking primers (Appendix A).

2.2.4 Sequence Analysis
Sequence annotation and translation were performed using Gene Runner software (version 3.05, Hastings Software Inc.). Sequence alignments were performed using the online pairwise BLAST (Basic Local Alignment Search Tool) (National Center for Biotechnology Information ( and the DNAMAN software (version 4.0, Lynnon BioSoft, Quebec). Fragment assemble was performed using the ContigExpress software, part of the Vector NTI suit of programs (version, Informax Inc. Bethesda, Maryland).

2.2.5 Isolation of Genomic DNA
Alfalfa genomic DNA was isolated from a pooling of mature and developing leaf tissue and purified using the Trebuchet gDNA extraction protocol (Bowley Lab, University of Guelph). Tissue (1 g) was ground by vortexing in a 50 ml centrifuge tube with ceramic beads and liquid nitrogen. Once the nitrogen had evaporated, 15 ml of extraction buffer (100 mM Tris pH 8.0, 50 mM EDTA pH 8.0 and 500 mM NaCl) was added and vortexed. To the tube, 1 ml of 20 % SDS was added, vortexed and the solution were incubated at 65°C for 10 min. To precipitate proteins and sugars, 5 ml of 5 M potassium acetate was added, vigorously mixed and incubated on ice for 20 min. Cell debris was pelleted by centrifugation at 25 000xg for 20 min. The supernatant was filtered through Miracloth (Calbiochem) into a 50 ml centrifuge tube containing 12.6 ml of ice cold isopropanol. Using a pipette, the cloudy interphase layer was removed and discarded. The tube was gently inverted and incubated at -20°C for 2 hrs. Using sealed glass pipette with a hook tip, the precipitated DNA was removed from solution, transferred to a 1.5 ml centrifuge tube and spun at 12 000xg for 2 min at 4°C. The supernatant was discarded and the pellet was washed 2x in cold 75% ethanol. The tube was inverted for 10 min and then vacuum dried for 10 min. The pellet was dissolved in resuspension buffer (1 M Tris and 0.5 M EDTA, pH 8.0) overnight at 4°C. To the tube, 20 ug/ml RNase A (1.4 ul of 10 ug/ul stock) was added and incubated at 37°C for 30 minutes. The DNA was precipitated by adding 75 ul of 3 M sodium acetate and 500 ul of cold isopropanol. As above, the DNA was pelleted, washed and resuspended in 100 ul of H2O and incubated at 40°C overnight.

2.2.6 Genomic Library Construction
A genomic library was built using the Universal GenomeWalker™ kit (Clontech, Cat# 1807-1, Palo Alto, California). The library was built according to the protocol included in the kit (available at Genomic DNA was digested in five separate library reactions, individually using EcoR V, Dra I, Pvu II and Ssp I. Genomic DNA adapter fragments containing unique primer sites were ligated onto the ends of the digested genomic DNA from each reaction. Genomic libraries were stored at -20°C until used.

2.2.7 Genomic Library Screening
Three primer sets were synthesized to PCR the genomic sequence from the library. The primers used for screening are listed in Appendix A and include MsHSF-5 through MsHSF-9. Primer design and PCR conditions were followed as according to the Universal GenomeWalker™ kit protocol. Primer selection was accomplished using the Primer3 web tool ( Using gene specific antisense primers and primers specific for the ligated adapters, nested PCR was performed on each library. PCR reactions were separated on an agarose gel to identify PCR reaction tubes with positive amplification. PCR fragments were cloned directly from the PCR reaction tube using the TOPO TA cloning kit (Invitrogen). The genomic DNA inserts were sequenced and analyzed for promoter consensus elements.

2.2.8 Digoxigenin (DIG) Probe Preparation
A 1011 bp DNA probe was generated from the entire MsHSFA4 open reading frame (ORF). MsHSFA4 plasmid was digested with Eco RI to release the insert and the ORF was isolated by gel purification on a 0.9 % TBE agarose gel run at 70 volts for 2 hrs. The ORF fragment was purified from the agarose gel by a freeze thaw extraction (Sambrook et al., 2000). The DIG PCR probe reaction was carried out using primers specific to the ORF producing a 1 kb fragment. A 1 ul sample of the probe reaction and 1 ul of a control reaction were separated on a 0.8 % agarose gel to observe a band shift indicating successful incorporation of the DIG labeled nucleotides.

2.2.9 Southern Blot Hybridization
Alfalfa genomic DNA (20 ug ) was digested overnight in four separate reactions of 200 ul each using Xba I, Hind III and Sac I respectively. The DNA in each reaction was precipitated (1/10th volume sodium acetate and 2.5 volumes of 100% ethanol) for 2 hrs, spun down at 12 000xg and re-suspended in 20 ul of deionized H2O. Samples were separated on a 1% agarose gel run at 15 volts for 14 hrs. The DNA was capillary blotted (in 20x SSC) from the gel onto a positively charged nylon membrane (Roche Diagnostics, Laval, Quebec cat # 1417240) overnight and fixed to the membrane by UV cross-linking. The membrane was rinsed briefly in deionized water to remove excess SSC and placed into a glass hybridization tube containing 30 ml of Easy Hyb (Roche cat # 1796895) solution without probe. The pre-hybridization continued for 3 hrs at 37°C, rotating in a hybridization oven. DIG labeled DNA probe was boiled for 3 min to denature the DNA, cooled on ice for 3 min and added to 30 ml of Easy Hyb solution. The pre-hybridization solution was removed from the tube and the Easy Hyb solution containing the probe was added and incubated at 37°C overnight in a rotating hybridization oven. Stringency washes and detection were preformed according to the DIG manual (Roche).
Prepared membranes were over-laid with CDP-Star chemiluminescent substrate solution and heat sealed between two pieces of clear acetate, as described in the provided Roche manual. The sealed membrane was exposed to Kodak Biomax film (Fisher Scientific, cat# 057281) for 10 and 20 min exposures and developed.

2.2.10 Fluorescent In Situ Hybridization
Scarified alfalfa seeds (50-60), cv. Marathon, were sealed in a Petri dish on a circular filter paper saturated with sterile H2O to germinate. The dish was incubated in the dark at 28°C for 24-36 hrs (some seeds take longer to germinate) to produce approximately a 1 cm root. The roots were harvested and transferred to glass vials on ice and the whole container was incubated at 4°C overnight. Roots were then transferred to a 1.5 ml microtube containing 1 ml of fixative (3:1 v/v 100% ethanol : glacial acetic acid) and incubated for 30 min, then were transferred to a Petri dish containing deionized H2O to rinse out the fixative. Root tips were excised using a scalpel (approximately the end 2-3 mm of the root) and transferred to a 1.5 ml microtube containing enzyme mix (3 % cellulase, Sigma cat# C-1794 and 1 % pectolyase, Sigma cat# P-3026) and incubated 37°C for 2 hrs. The enzyme mix was removed and the root tips were gently washed with H2O. Using a glass pipette four root tips were transferred to a glass slide (Fisher cat# 12-550-12) and any excess water was removed capillary action using a piece of paper towel. The root tips were gently broken up by tapping with needle point forceps. A drop of fixative was placed over the crushed tips, flamed to remove excess fixative and air dried. Slides could then be stored at -80°C for up to 3 months.

A biotin labeled probe was prepared from total plasmid containing the MsHSFA4 cDNA by nick translation using the BioNick Labelling System (Invitrogen, cat # 18247-015). The synthesis reaction and quantification were completed as according to protocol included with the kit. The probe was mixed with salmon sperm DNA and aliquoted into tubes (5 ul of 200 ng/ul).

The slides were thawed and allowed to air dry. To each slide 200 ul of formamide denaturing solution (Appendix B) was applied, covered with a glass cover slip and incubated at 80°C for 3 min. The cover slips were then removed and slides were transferred to a glass rack, placed in 70 % ethanol for 5 min at 20°C, 95 % ethanol for 5 min at -20°C, 100 % ethanol for 5 min at -20°C, then allowed to completely air dry for 20 min. The slides were transferred to a humidity and temperature controlled incubation chamber moistened with 2x SSC in the bottom of the chamber. Slides were arranged so that they were not touching and were raised above the SSC. To each slide 40-50 ul of probe solution were added and covered with a glass cover slip. Slides were incubated at 80°C for 20 min to denature the probe and the chromosomes, then hybridized at 35°C for 18 hrs.

Following hybridization the cover slips were removed with 2x SSC and approximately 150 ul of blocking solution (4xSSC, 0.2 % Tween-20 and 3 % BSA) were applied. The slides were covered with small squares of cellophane and incubated in the humidity chamber at room temperature for 10 min. The blocking solution was drained off,150 ul of Fluorescent Avidin DCS (Vector Laboratories, cat # A-2011) solution (8 ul of Fluorescent Avidin DCS in 1 ml of blocking solution) of were added, again covered with small squares of cellophane, and incubated at 37°C for 40 min. Slides were then washed in 2x SSC at 42°C for 10 min and then washed in 0.5x SSC and 0.2 % Tween-20 (Sigma) at 42°C for 10 min. The slides were blocked for a second time by adding 150 ul of blocking solution and incubating at room temperature for 10 min. The blocking solution was drained and 150 ul of biotinylated anti-Avidin D (Vector Laboratories cat # BA-0300) solution (20 ul in 1 ml of blocking) were added and incubated in a humidity chamber at 37°C for 30 min. Slides were then washed in 2x SSC at 37°C for 15 min. The excess was drained off and 150 ul of Fluorescent Avidin DCS solution were added to each slide and incubated at 37°C for 40 min. Slides were then washed in 2x SSC at room temperature for 10 min. The slides were drained and mounted with a glass cover slip over a drop of anti-fade solution (Invitrogen). The slides were incubated at 4°C in the dark overnight and viewed the following day.

Images were captured at 1000 x magnification under oil immersion using a digital camera connected to a Leitz phase contrast fluorescent microscope. Images of chromosome spreads were captured using the Northern Eclipse Image Analysis Software (Empix Imaging, Mississauga, Ontario). Individual chromosomes were karyotyped using the MicroMeasure software (version 3.3, Aaron Reeves & Jim Tear, Colorado State Univeristy) and compared with alfalfa chromosome analyzes from the literature (Bauchan and Hossain, 1998; Bauchan and Hossain, 1997).

2.2.11 RT-PCR
Total RNA was isolated from a pooling of mature and developing leaf tissue using the RNAqueous Kit (Ambion, Austin Texas, cat # 1912). The extraction was performed as according to the protocol supplied with the kit. Total RNA was quantified using the RiboGreen fluorescent RNA stain (Molecular Probes) and detected with a Shimazo 4200 Flurometer. A 1 ul aliquot of each RNA sample was added to 2 ml of 1x TE buffer (RNase free) and 5 ul of stock Ribogreen stain in a four translucent wall polypropylene cuvette. The RNA and Ribogreen stain were incubated in the dark for 5 min to allow the stain to bind the RNA and the fluorescence of each sample was measured with an excitation wavelength of 480 nm and reading wavelength of 520 nm. Standard curves were generated from RNA standards, provided with the Ribogreen stain. RNA quality was assessed on 1.2 % denaturing agarose gel containing 5 % formaldehyde.

RNA samples were treated with RNase free DNase I following the protocol supplied with enzyme (Promega, cat # 610A). Removal of excess enzyme and protein was accomplished using a DNase inactivation reagent (Ambion, cat# 8173G) and following the supplied protocol. Converting the RNA to cDNA was accomplished using SuperScript II reverse transcriptase (Invitrogen, cat# 18064-014) and the supplied protocol with the addition of 0.5 ul of RNase inhibitor (Promega, cat# N211A) per reaction. The PCR was accomplished using Platinum taq DNA polymerase (Invitrogen, cat # 10966-026) following the supplied protocol and using the following cycling conditions; denaturing 94°C for 45 sec, annealing at 58°C for 45 sec, extension at 72°C for 1.5 min and 44 cycles. PCR products were analyzed on a 0.8 % agarose gel.

2.3 Results

2.3.1 cDNA Sequence
The cDNA insert from the purified positive phage clones was sequenced from both the 5’ and 3’ ends using the T3 and T7 universal primer sites built into the phagemid vector (for primer sequence see appendix A). The sequencing revealed that a cloned cDNA of approximately 2 kb was isolated from the alfalfa bud cDNA library. BLASTx nucleotide analysis revealed that this clone exhibited a high degree of similarity to known HSFs, specifically, Phaseolus acutifolius PaHSFA4 (GenBank accession number AY052627), Nicotiana tabacum NtHSFA4, formerly known as NtHSF2 (GenBank accession number AB014484), and Arabidopsis thaliana AtHSFA4a, formerly known as AtHSF21 (GenBank accession number U68561 (Figure 2.1). Parsimony analysis established the alfalfa amino acid sequence as belonging to the plant HSF subclass A4 and it was thus named MsHSFA4 (GenBank accession number AF235958) (for complete sequence and translation see appendix D).

Comparison of MsHSFA4 to the PaHSFA4 nucleotide sequence revealed a high degree of similarity in both the overall cDNA sequence (62% homologous) and the region encoding DBD (87% homologous) (Figure 2.1, B). In contrast NtHSFA4 and AtHSFA4 possessed less than 50% similarity in the overall cDNA sequence yet they both possess 80% similarity in the region encoding the DBD (Figure 2.1, B). MsHSFA4 and PaHSFA4 both contain large 5’ UTRs (over 500 bp) as compared with NtHSFA4 and AtHSFA4 that have average size 5’ UTRs (under 300 bp). Although the nucleic acid analysis of the ORFs revealed varying degrees of similarity, the overall size of the coding regions was extremely close in length (Figure 2.1, A). 5’ UTR
The large size of the MsHSFA4 5’ UTR was found to be atypical of other HSFs recorded in the literature. To confirm that this 5’UTR was not an anomaly of the library construction process, RT-PCR was employed to verify this UTR existed in the expressed mRNA from shoot tissue. Primer pairs specific to the 5’ UTR and primers that span the 5’ UTR and the open reading frame were used to observe the integrity of the full mRNA and determine if there was any temperature responsive transcription or alternative splicing. Analysis of the RT-PCR revealed that all expected bands were present and had the appropriate relative sizes (Figure 2.2). However, the 5’ UTR bands were approximately 100-150 bp less than the predicted (predicted sizes listed in Figure 2.2, A), indicating that a sizeable piece was spliced out. Therefore the large 5’ UTR observed in the MsHSFA4 sequence is present in mature transcripts but, modified. There was no difference in banding patterns among control and heat shock tissues suggesting the large 5’ UTR or the full sequence is consistently modified regardless of temperature.

2.3.2 Amino Acid Sequence
Translation of the MsHSFA4 cDNA revealed a protein of 402 amino acids with a molecular weight of 46.2 kilodaltons (kDa) (GenBank accession number AAF37579). The overall protein sequence has similarities to PaHSFA4 , NtHSFA4 and AtHSFA4 of 76%, 66% and 62% respectively (Figure 2.3). However in the DBD, similarity was 89 to 90% among all three sequences. MsHSFA4 contains several conserved regions located C-terminally to the DBD including HR A and HR B, a nuclear localization signal and two AHA motifs located in the CTAD. All of the conserved regions aligned almost perfectly to aforementioned HSFs except for the AHA motifs.

2.3.3 Genomic Sequence
Isolation and sequencing of the MsHSFA4 genomic DNA revealed this gene had two exons of 802 bp and 1159 bp that subtend a single intron of 105 bp (Figure 2.4). This genomic clone was re-sequenced 3 times for accuracy and submitted to GenBank (accession number AF494082) (for complete sequence and translation see appendix E). Limitations in the library screening procedure restricted the amount of up-stream sequence identified to a region of 278 bp ahead of the 5’ UTR. Within this region a number of promoter elements were identified, specifically, a putative TATA box, a heat shock element (HSE) and several HSE head and tail modules (Figure 2.4). Analysis of current sequence databases revealed that there is only sequence data available for two other plant class A4 HSFs specifically, Lotus japaonicus (see Appendix C for phylogeny relatedness to alfalfa) and Arabidopsis thaliana. The promoter region of both of these sequences exhibits less than 50% homology as compared to the MsHSFA4 promoter region, but contained a number of conserved motifs including a putative TATA box, an active HSE motif and active HSE head and tail motifs (figure 2.4, C).

2.3.4 Genomic Localization
Southern blot analysis using a probe to the MsHSFA4 ORF revealed a single strong band and a possible faint second band. The result suggests that MsHSFA4 is either a single copy gene or double copy gene (Figure 2.5). Chromosome spreads exhibited the expected 32 visible chromosomes (8 homologues), confirming the source alfalfa plants as tetraploid alfalfa. Chromosomal localization of MsHSFA4, through fluorescent in situ hybridization revealed that this gene exists as a single copy gene with four homologous copies in a tetraploid genome. Based on repeated observations of arm ratio and relative average total chromosome length, this locus exists on either chromosome 6 or 7. Among the four homologous chromosomes MsHSFA4 resides at one of two genetic loci (Figure 2.6). The first locus is found near the telomere and the second locus is found near the centromere. Comparisons to chromosomal locations of known class A4 revealed that in Lotus japonicus (2n=2x=12), its class A4 HSF was localized to chromosome 4 and in Arabidopsis thaliana (2n=2x=10) its class A4 HSF was also localized to chromosome 4. This indicates a possible conservation of HSFA4 gene location.

2.4 Discussion

MsHSFA4 is a single copy gene that exhibits high structural similarities to known HSFs. This conservation is most prominent at the protein level indicating that its function has been conserved within plants. At the DNA level the sequence conservation decreases among the three species analyzed with evolutionary relatedness to alfalfa. This is exemplified in comparing the relatedness of the Phaseolus and Arabidopsis HSFs to the alfalfa homologue. Phaseolus which bears the greatest similarity is found within the same family as alfalfa, Fabaceae, whereas Arabidopsis, which exhibited the least similarity of the three sequences surveyed, is found in the Brassicaceae or mustard family. Their closest evolutionary branching point is at the level of class, Dicotyledons (for species classification see appendix C).
As is observed in the Phaseolus HSF, the alfalfa sequence has an abnormally large 5’ UTR, which was not present in the Nicotiana and Arabidopsis cDNAs. This indicates that class A4 HSFs in the Fabaceae family may have large 5’ UTR’s. Currently a number of HSFs have been isolated for species in the Fabaceae family, including Glycine max, Pisum sativum and Phaseolus acutifolius; however, reliable 5’UTR data only exists for alfalfa and Phaseolus A4 family members. The function of the large 5’ UTR is unknown but, the presence of this extended 5’ UTR, in conjunction with the observation that a portion of the sequence is not present in the mature transcript, indicates that it has a regulatory effect, specifically, repressing its translation (see Chapter 3, 3.3.1).

Alfalfa chromosomes are small and relatively similar in size which proves extremely difficult to identify different chromosomes using an in situ hybridization procedure. C-banding which is regularly used to distinguish alfalfa chromosomes, cannot be used with in situ hybridization. Relative length and arm ratios were the only tools that could be employed to discern between chromosomes. Analysis of a number of chromosome spreads showed that MsHSFA4 was most likely located on either chromosome 6 or 7. Localization of MsHASFA4 has found two loci for this single copy gene. In an auto-tetraploid, one expects to find a single copy gene at one locus among the four homologous chromosomes. However, alfalfa exhibits one locus near the telomere and the second locus near the centromere. Cytologically, this type of locus orientation is usually indicative of a large chromosomal inversion event. However, two homologues have the telomere locus and the other two homologues have the centromere locus, a 2:2 locus ratio, as opposed to a 1:3 locus ratio. This suggests that the inversion was restricted to the chromosome arm and the inversion event occurred before alfalfa’s fusion from diploid to tetraploid. Furthermore, no genotypes were detected that had other combinations of the loci. In an auto-tetraploid one would expect random assortment resulting in a range of genotypes having zero, one, two, three or four copies of the chromosome with the telomere locus and the same for the centromere locus. Only genotypes with two copies of the telomere and two copies of the centromere locus were identified. This indicates that sections of the alfalfa genome act as a diploid and the species is an intermediate between an allo-tetraploid and an auto-tetraploid.

2.5 Conclusion
A new HSF homologue, designated MsHSFA4 was isolated from alfalfa and is highly homologous to other plant class A4 HSFs. This homology was notable at the DNA level, but was profoundly apparent at the protein level. Structurally, MsHSFA4 displayed high conservation in the potential functional domains including the DBD, OD and the CTD. MsHSFA4 is a single copy gene located on either chromosome 6 or 7 at one of two loci.