This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
A, a Hedera helix shoot tip showing the first unfolded leaf (L3), and the younger leaf (L2), which was used as the young leaf specimen without petiole. Leaf #4 was removed at the time of shoot tip sample collection in the greenhouse. B, Longitudinal view of hand-section of the shoot tip showing the shoot apex (SA), the enclosing leaf #1 (L1), leaf #2 (L2) and leaf #3 (L3). Leaf #2 was cut through mid-veins of the blade; and leaf #3, through the center of the petiole. The typical shoot apex sample included small developing young leaves, leaf primordia and the apical dome, which were enclosed by the Leaf #1. Leaf #1 was not included in the shoot apex sample.
Figure 2-1. Clonal lineage of sample plants propagated by stem cuttings.
The numbers in the circles are the quantities of plants produced by stem cutting, while the numbers in the call-out boxes are the quantities of the stock plants from which the stem cuttings were obtained. A: Starting plant stock, 10 plants obtained locally. B: The first batch of plants produced by stem cuttings from each of the starting stock. C: The second batch consisting of 135 plants, which were obtained from 13 of 27 plants propagated from a single plant. D: Out of 135 plants in the second batch, 100 were transferred to BGSU and 60 new shoots were re-grown. E: Among 60 plants with newly grown shoots, 16 were used to produce the third batch of 193 plants. F: Another 40 plants were used for stem cutting of the fourth batch consisting of 574 plants. G: More than 400 plants, which were propagated from a single stock plant, were maintained for sampling.
Figure 2-3. A schematic diagram showing preparation of fluorescently labeled cDNA targets.
Total RNA was isolated from each sample tissue. Using total RNA as template, amino-modified 1st-stand cDNA was synthesized. The amino-modified cDNA solution from each sample was divided in half. For dye-swap, shoot apex cDNA target was labeled with Alexa Fluor? 555, and young leaf cDNA with 647 in a target set. In the other target set, dyes were coupled reversely; i.e., shoot apex cDNA target was labeled with Alexa Fluor? 647, and young leaf cDNA with 555.
Figure 2-4. Die-Technology hybridization chamber model DT-1001.
Grooves and reservoirs are for chamber buffer to keep moisture in the chamber during incubation for prehybridization or hybridization. Capacity of each groove and reservoir are 50 and 100 ?l, respectively, holding 300 ?l total. (Labels added to a downloaded picture from http://www.die-technology.com/images/p_DT-1001_big.jpg)
Figure 2-5. A schematic diagram showing hybridization in dye-swap design.
Array probes on each array slide were hybridized with cDNA targets that were opositely labeled. In Array 1, shoot apex cDNA target labeled with Alexa Fluor? 555 and young leaf cDNA target labeled with Alexa Fluor? 647 were co-hybridized, while in Array 2, shoot apex cDNA target labeled with Alexa Fluor? 647 and young leaf cDNA target labeled with Alexa Fluor? 555 were co-hybridized.
Part 2. Materials and Methods
2.1 Plant Material
2.1.1 Starting Plant Stock
Ten variegated ivy (Hedera helix L. cv. Goldheart) plants, which were assured by the vendor to have been propagated from single plant, were obtained locally (Chuck Hafner's Farmers Market, Syracuse, NY) and transferred in plastic pots (4" diameter) containing a 1:1:1 mixture of perlite, vermiculite and MiracleGro? Professional Potting Mix (Scotts Co., Marysville, OH). The potted plants were maintained under a greenhouse bench, where they were partially shaded, in the greenhouse at the State University of New York College of Environmental Science and Forestry (SUNY ESF), Syracuse, New York. About three months after transplanting, when root systems were stabilized and shoots resumed growing vigorously, the main shoots of the plants were cut off and new shoots were grown from the lateral buds at the first, second or third node from the ground. These new shoots were used for propagation by stem cutting (Figure 2-1, A).
2.1.2 Propagation by Stem Cutting
After one growing season, stem cuttings were prepared separately from each plant. Stem segments of 10-15 cm in length, containing 3-5 nodes, were cut using sharp razor blades. After removing the leaves at the lowest node, the base of the cuttings were soaked in an indol-3-butyric acid (IBA, # 57310; Sigma-Aldrich, St. Louis, MO) solution at 1000 ppm for one minute, powdered with Captan? 50% WP (Bonide Products, Inc. Oriskany, NY), and then immediately inserted into the holes pre-made in a 1:1 mixture of perlite and vermiculite (Horticultural Grade; Conrad Fafard, Inc., Agawam, MA). Stem cuttings were kept for about 1? months until they produced new root and shoot systems under humid conditions maintained by a mist system. More than 95% of the total cuttings rooted successfully and 13-27 new plants per stock were produced for further propagation (Figure 2-1, B). Plants were grown in the greenhouse at SUNY ESF for about 10 weeks until they produced 15-20 nodes. They were watered regularly and fertilized with MiracleGro? All Purpose Plant Food (Scotts Co., Marysville, OH) as needed following the manufacturer's recommendation. To control aphids and white flies, Ortho? Isotox? Insect Killer (Scotts Co., Marysville, OH) was sprayed periodically following the manufacturer's instruction.
Using the newly established ivy plant stock, the second stem cutting propagation was carried out in the same way as described previously. To ensure genetic homogeneity of sample plants, only the batch that was in the largest number and relatively uniform in the size was selected. A total of 135 plants were obtained by the second stem cutting from 13 of the 27 plants propagated from single plants from the initial batch (Figure 2-1, C).
2.1.3 Re-establishment of Plant Batches
About 100 plants were transferred to Bowling Green State University (BGSU), Bowling Green, Ohio. Shoots were cut back and new shoots were grown from about 60 plants (Figure 2-1, D), which were previously propagated by stem cuttings from a single plant. A new batch of 193 plants was established by stem cuttings from 16 plants (Figure 2-1, E). Stem cutting was carried out in the same way as described above, except for the cutting treatment, where the cuttings were treated with Hormex? Rooting Powder No. 1 (Brooker Chemical, Chatsworth, CA) containing 0.1% indol-3-butyric acid (IBA) in talc powder.
Another batch of 574 plants was propagated from another 40 H. helix cv. Goldheart plants (Figure 2-1, E). Cuttings were treated with Hormex? Rooting Powder No.1 (Brooker Chemical, Chatsworth, CA), and rooted on OASIS? WEDGE? System (Cat. #5644; Smithers-Oasis U.S.A., Kent, OH), a rooting medium formulated to assure a high soil moisture content. A month later, the plants which rooted in the wedge media were transferred to 5?" square plastic pots containing a fertilized soil mixture. Plants were watered and fertilized as needed and pesticides were applied as needed to control white flies and aphids. More than 400 plants, which originated from single plant, were maintained in the greenhouse throughout the period of sampling.
2.2 Sample Tissue Preparation
2.2.1 Tissue Collection
Samples of shoot apices and young leaves always were collected from actively growing healthy shoots. And, after sampling, the shoots were cut back leaving 2-3 nodes, at which a new shoot was grown for the next sampling. A shoot tip containing a shoot apex and a young leaf subtending the shoot apex was excised from each of 250 plants (Figure 2-2, A) and immediately flash-frozen in liquid nitrogen. The jar containing liquid nitrogen and shoot tip specimens was capped loosely and transferred to a Styrofoam box containing dry ice. After all of the liquid nitrogen evaporated completely, the jar containing the shoot tip samples was capped tightly and transferred to the laboratory for further dissection. The shoot tips were stored in a -80?C freezer until they were dissected into apex and leaf specimens.
For dissecting, the shoot tips were kept submerged in liquid nitrogen to prevent formation of frost. Under a dissecting microscope, a shoot tip was placed on an aluminum block chilled by dry ice-95% ethanol bath, and isolated the young leaves from the shoot apex. Starting from the largest one, three young leaves were excised at the leaf base without petioles using a pair of fine-pointed forceps (Inox #5; Dumont S.A., Switzerland) and a scalpel with blade #11 (Feather Safety Razor Co., Osaka, Japan). The excised leaf samples were collected separately in insulated plastic jars containing liquid nitrogen according to their developmental stages, as follows (Figure 2-2, B):
1. Young leaf #3: the smallest completely unfolded leaf with leaf blade larger than 1 cm in length.
2. Young leaf #2: partially or completely folded, but not covering the shoot apex; used for total RNA extraction in this experiment.
3. Young leaf #1: completely folded and enclosing the shoot apex.
After collecting the young leaf specimens, the apex specimens were excised and collected in a separate jar containing liquid nitrogen. Each apex specimen typically included the shoot apical meristem, leaf primordia and the first two emerging leaves; the first leaf had its concave adaxial surface adjacent to the primordium, and the second leaf, at the opposite side of the first leaf, had a leaf blade edge bending over and covering the apical dome. A total of 230 shoot apices and 100-115 young leaves were collected by dissection. The tissue samples were stored in a -80?C freezer until isolation of total RNA.
All of the instruments and containers used for collection and dissection were RNase-free and the entire sampling procedure was carried out in an RNase-free environment, whenever possible. The instruments and containers were cleaned with RNaseAway? (Cat. #10328011; Invitrogen, Carlsbad, CA) or RNaseZap? (Cat. #9780; Applied Biosystems/Ambion, Austin, TX), rinsed with diethyl pyrocarbonate (DEPC; Cat. #D5758; Sigma-Aldrich, St. Louis, MO)-treated RNase-free water (Sambrook and Russell, 2001), and subsequently with 100% ethanol. The dust-free bench top was sprayed with one of the RNase decontamination agents and wiped with Kimtech Science? KimWipes? Delicate Task Wipers (Cat. # 34256; Kimtech Science, Roswell, GA). After wiping with the wipes moistened with DEPC-treated RNase-free water, the surface was wipe-dried using the same type of wipes.
2.2.2 Grinding Tissue Samples
For the extraction of total RNA, the shoot apices and young leaf samples were ground into fine powder. Frozen samples were ground in liquid nitrogen in RNase-free 15 ml BD Falcon? polystyrene conical tubes (Cat. #352095, BD Biosciences, San Jose, CA) using an RNase-free Kontes? Chlorotrifluoroethylene (CTFE)/Stainless Steel Pellet Pestle? for 1.5 ml Tube (Cat. # 749515-0000; Kimble-Kontes, Vineland, NJ) attached to a cordless rotary tool (Dremel? Minimite 750, 4.8V; Robert Bosch Tool Corp., Racine, WI). The lower portion of the tube containing tissue sample was kept in liquid nitrogen, and the sample inside was slightly covered with liquid nitrogen to eliminate frictional heat between the tissue, pestle and tube wall. When the tissue was ground to fine powder suitable for the extraction of total RNA, it was pale green in color and dispersed readily in slowly swirled liquid nitrogen, forming aggregates, which eventually settled on the bottom of the tube as a uniform layer of talc-like powder. After grinding, the tube containing liquid nitrogen and tissue powder was loosely capped and placed on dry ice at a 45 degree angle to evaporate the liquid nitrogen without excessive boiling and splashing. When all of the liquid nitrogen evaporated, the tube was capped tightly and kept at -80?C until RNA isolation.
2.3 Total RNA Extraction and Purification
Total RNA was extracted from the ground tissue samples of shoot apex and young leaf sample. To maximize the yield of isolated total RNA at the highest purity possible, total RNA was extracted with Trizol Reagent? (Cat.#155960; Invitrogen, Carlsbad, CA; Chomczynski and Mackey, 1995). The RNA-containing aqueous phase was separated using Phase Lock Gel? (PLG, Heavy; Cat. # 95515404-5 or # 95515407-0; Eppendorf North America, Westbury, NY) and purified using RNeasy? Midi columns (Cat. #75142; Qiagen, Valencia, CA) in combination with RNase-free DNase (Cat. #79254; Qiagen, Valencia, CA) treatments.
The frozen sample powder of shoot apices and young leaves were transferred into 15 ml or 50 ml BD Falcon? polystyrene conical tubes (Cat. #352095 or #352073, BD Biosciences, San Jose, CA), which were pre-weighed and pre-chilled in liquid nitrogen. After the approximate volume of the sample and the net tissue weights were determined, approximately 10X sample volume of Trizol Reagent? was added to each tube (Table 2-1). The reagent stored at 4?C, was warmed to 35-40?C in a water bath before being added to prevent freezing when it was added to the frozen tissue. Immediately after adding the reagent, the mixture was homogenized thoroughly using an RNase-free Kontes? Chlorotrifluoroethylene (CTFE)/Stainless Steel Pellet Pestle? For 0.5 ml Tube (Cat. #749515-0500; Kimble-Kontes, Vineland, NJ) attached to a Pellet Pestle Cordless Motor (#749540-0000; Kimble-Kontes, Vineland, NJ), and then incubated at room temperature for 6 minutes with vortexing every 2 minutes.
Chloroform (0.2 volumes relative to Trizol Reagent?) was added to each tube containing tissue-Trizol? homogenate, and the contents were mixed thoroughly to form a homogeneous suspension by vortexing. The tissue-Trizol? homogenate/chloroform mixture was transferred into pre-spun PLG tubes (Heavy; Cat. # 95515404-5 or # 95515407-0; Eppendorf North America, Westbury, NY) and mixed thoroughly by shaking vigorously. To separate the organic and the aqueous phases, 1.5 ml aliquots of Trizol homogenate/chloroform mixture were transferred into each of pre-spun 2.0 ml PLG tubes and mixed thoroughly by shaking vigorously. The PLG tubes were centrifuged at maximum speed (13,200 rpm) on a microcentrifuge (Model 5415D; Eppendorf North America, Westbury, NY) for 15 minutes to separate the phases. To keep the temperature of the PLG tubes low, dry ice was placed around and on the centrifuge. Immediately after centrifugation, the RNA-containing aqueous phase was pipetted off and pooled in a fresh RNase-free 15 ml BD Falcon? polystyrene conical tube (Cat. #352095, BD Biosciences, San Jose, CA) for each sample.
To purify total RNA from the pooled aqueous phase, an equal volume of 70% ethanol was added slowly to the retrieved aqueous phase and mixed thoroughly by inverting the tube repeatedly. The first 4 ml aliquot of the mixture was pipetted into a RNA-binding column placed in a collection tube, which was supplied in the RNeasy? Midi kit (Cat. #75142; Qiagen, Valencia, CA). The column was centrifuged for 5 minutes at maximum speed (4750 rpm ? 2%) on a bench-top centrifuge with a swing-bucket rotor (Centrific? Model 225; Fisher Scientific, Springfield, NJ). After the centrifugation, flow-through in the collection tube was discarded. Centrifugation was repeated using the same RNA-binding column for the rest of the mixture.
DNase I digestion was carried out using the RNase Free DNase Set (Cat. #79254; Qiagen, Valencia, CA) following the manufacturer's protocol. Briefly, the total RNA-bound RNeasy? silica-gel membrane in the binding column was washed with 2 ml RW1 buffer, which was included in the RNeasy? Midi kit, by centrifugation for 5 minutes at maximum speed. After applying 160 ?l of the incubation mixture containing approximately 55 Kunitz units of DNase I directly onto the RNeasy? silica-gel membrane, the reaction was incubated on the bench top at room temperature for 15 minutes. After DNase I digestion, the RNeasy? membrane was washed by two successive loadings of 2.5 ml RPE buffer supplied in the RNeasy? Midi kit and centrifugation for 2 minutes each at the maximum speed using the same centrifuge as in the RNA binding step described above. To dry the RNeasy? silica-gel membrane, the tubes were centrifuged for additional 3 minutes at maximum speed following the second centrifugation.
Total RNA was eluted by applying two 100 ?l aliquots of non-DPEC treated RNase-free water supplied in the RNeasy? Midi Kit. After adding each aliquot of eluent, tubes were incubated for 5 minutes at room temperature and then centrifuged at maximum speed for 3 minutes. Using water in the second elution results in a higher yield of total RNA while using the first eluate as an eluent increases the concentration of total RNA with lower yields (Qiagen, 2001). To obtain a higher total RNA yield, a second elution step was performed using another volume of RNase-free water, rather than using the first eluate as the eluent to increase the concentration of total RNA.
2.3.3 Storage, Quantitation and Determination of Quality of Total RNA
The total RNA solutions were transferred to 1.5 ml microcentrifuge tubes with screw caps and kept frozen in a -80?C freezer, until needed. The quantity and purity of total RNA purified with RNeasy? Midi columns were assessed by spectrophotometry using 100 ?l quartz cuvettes in a spectrophotometer (Model DU-600; Beckman, Fullerton, CA). For measurements of concentrations, the total RNA eluates were diluted in deionized water and the absorbance was measured at 260 nm with 320 nm background correction. RNA concentration was calculated by the equation:
(EQ. 2-1)(A260 ? A320) ? 40 mg/ml
which is based on an extinction coefficient calculated for RNA in water (Qiagen, 2001a; Qiagen, 2001b). As the relationship between the absorbance and concentration is reliable only when the absorbance readings are in the range between 0.15-1 (or possibly 1.2), dilutions were made so that the absorbance readings of the diluted sample was within this range.
For determination of the purity of RNA, the total RNA samples were diluted in 0.1X TE (10 mM Tris-HCl, pH 7.5), and the spectrophotometer was calibrated with the same solution (Wilfinger et al., 1997). Absorbance was measured at 260 and 280 nm with 320 nm background correction. The ratio (A260 - A320)/(A280 - A320) was calculated to estimate the purity of RNA with respect to contaminating protein that absorb in the UV. The programmed routine in the equipment did not calculated the ratio (A260 - A320)/(A230 - A320), which is an estimation of carbohydrate contaminants. The integrity of total RNA were verified by the ratio between 25S and 18S ribosomal RNA on a denaturing agarose gel electrophoresis (Imbeaud et al., 2005; Kleber and Kehr, 2006), which was run using the protocol and reagents described in the NorthernMax? Kit manual (Cat. #1940; Applied Biosystems/Ambion, Austin, TX) and SYBR? Green II RNA gel stain (Cat. #S-7564; Invitrogen, Carlsbad, CA).
2.4 Labeling cDNA Targets with Fluorescent Dyes
To incorporate amino-modified nucleotides into cDNA targets, the first-strand cDNA was synthesized by reverse transcriptase PCR (RT-PCR) using each total RNA sample as templates. Reactions were carried out using SuperScript III reverse transcriptase and poly-d(T) primers provided in the SuperScript? Indirect cDNA labeling System (Cat. #L1014-02; Invitrogen, Carlsbad, CA) and by following instructions in the accompanied manual (Invitrogen Life Technologies, 2004).
2.4.1 Synthesis of Amino-Modified cDNA
As the first step to prepare fluorescent dye-coupled cDNA targets, two amino-modified nucleotides, an aminoallyl-modified nucleotide and an aminohexyl-modified nucleotide, were incorporated with other dNTPs using SuperScript? III Reverse Transcriptase (included in the kit) during the cDNA synthesis reaction.
Equal amounts of total RNA from shoot apices or young leaf samples were used as RNA templates in each reaction. Since the concentrations of the eluted total RNA of shoot apices and young leaves were different, the volumes of total RNA added to the reactions were adjusted to balance the amount of RNA template between two samples. Template/primer mixture tubes were prepared as in Table 2-2.
Four reaction tubes were prepared for each sample, and a control reaction, in which an RNA ladder was used as template, was included to determine the efficiency of the labeling procedure. To denature RNA templates, the template/primer mixture tubes were incubated at 70?C for 5 minutes, and then placed on ice for at least 1 minute for primer annealing. To continue with first-stand cDNA synthesis reaction, the reaction mixture including amino-modified nucleotides (Table 2-3) was added to each template/primer mixture tube and mixed.
Tubes containing the reaction mixture were incubated at 46?C for 3 hours using a thermocycler (Mastercycler? Gradient; Eppendorf N.A., Westbury, NY) to synthesize first-strand cDNA with amino-modified nucleotides incorporated. Immediately after stopping the cDNA synthesis reaction by heating the tubes at 94?C for 2-3 minutes, RNA templates were hydrolyzed by adding 15 ?l of 1 N NaOH and incubating for 10 minutes at 70?C. Subsequently, the reaction mixture was neutralized by adding 15 ?l of 1 N HCl.
2.4.2 Purification of Amino-Modified cDNA
The amino-modified cDNA was purified to remove unincorporated dNTPs and hydrolyzed RNA templates using a QIAquick PCR Purification Kit (Cat. #28104; Qiagen, Valencia, CA) by following the procedures in the accompanying manual with some modifications. Instead of using EB buffer in the kit for elution, the amino-modified cDNA was eluted in the dye-coupling buffer (0.1 M sodium bicarbonate, pH 9.0), which was the buffer used in the subsequent labeling steps. The quality of purified amino-modified cDNA was assessed by agarose gel electrophoresis and ethidium bromide staining.
2.4.3 Concentrating Amino-Modified cDNA
The purified amino-modified cDNA solutions were concentrated using Microcon? YM-30 Centrifugal Filter Units (Cat. #42410; Millipore, Billerica, MA) to accommodate the whole amount of cDNA in the labeling reaction. For each tissue sample, two aliquots of 100 ?l amino-modified cDNA solution were transferred into two separate filter units. To obtain a final volume of 5 ?l, the filter units placed in the collection tubes were centrifuged for 6 minutes at 13,200 rpm on a microcentrifuge (Model 5415D, Eppendorf North America, Westbury, NY). When the recovered volume was less than 5 ?l, coupling buffer (0.1 M NaHCO3, pH 9.0) was added to make the final volume 5 ?l before the conjugation reaction with fluorescent dye.
2.4.4 Labeling Amino-Modified cDNA by Dye Conjugation
Subsequent to synthesis, purification and concentration, the amino-modified cDNA's were coupled with mono-functional forms of fluorescent dyes for labeling the target cDNA. Alexa Fluor? 555 and Alexa Fluor? 647 Reactive Dye Decapacks (Cat. #A-32755; Invitrogen, Carlsbad, CA) were used for labeling. For dye-swap (dye-flip or fluor-flip), two sets of fluorescently labeled cDNA targets were prepared (Table 2-4). One set consisted of shoot apex cDNA labeled with Alexa Fluor? 555 and young leaf cDNA labeled with Alexa Fluor? 647. The other set consisted of the same cDNA targets, but labeled with the opposite fluors (Figure 2-3).
To couple fluorescent dye to the amino-modified first-stand cDNA, dyes were prepared by adding 2 ?l of moisture-free DMSO directly to each of 4 dye vials; two containing Alexa Fluor? 555, and the other two containing Alexa Fluor? 647. The whole content of each dye vial was pipetted into each tube containing amino-modified cDNA target, such that one of two cDNA targets from each sample tissue was coupled with Alexa Fluor? 555 and the other with 647. To bring the final reaction volume to 10 ?l, 3 ?l of coupling buffer (0.1 M NaHCO3, pH 9.0) was added and mixed. The contents in the tube were mixed well and incubated at room temperature in the dark for 1 hour and 20 minutes.
2.4.5 Purification, Quantitation and Concentration of Labeled cDNA Targets
The fluorescently labeled cDNA targets were purified to remove any un-reacted dye using a QIAquick PCR Purification Kit (Cat. #28104; Qiagen, Valencia, CA). To calculate the total amounts of the amino-modified and the fluorescently labeled cDNA, respectively, in the purified target samples, absorbance were measured at 260 mn, 320 nm, 550 nm, 650 mn and 750 nm using a 'Multiple Wavelength Mode' in a Beckman DU-600 spectrophotometer (Beckman, Fullerton, CA). The labeled cDNA elutes were diluted in deionized water and measured in quartz cuvettes.
Total amounts of amino-modified cDNA were determined using the following formula:
(EQ. 2-2)Amino-modified cDNA (ng) = (A260?A320) ? 37 ng/?l ? 90 ?l (elution volume)
The amounts of fluorescently labeled dyes were calculated using the following formulas:
(EQ. 2-3)Alexa Fluor? 555 (pmole) = (A550?A650)/0.15 ? 90 ?l (elution volume)
(EQ. 2-4)Alexa Fluor? 647 (pmole) = (A650?A750)/0.24 ? 90 ?l (elution volume)
Labeling efficiency was assessed by comparing the yield of fluorescently labeled cDNA to the total yield of cDNA as described in the accompanied manual in the cDNA labeling system (Invitrogen Life Technology, 2004). After the spectrophotometry for labeling efficiency assessment, the purified labeled cDNA was concentrated using a Microcon? YM-30 Centrifugal Filter Unit (Cat. #42410; Millipore, Billerica, MA) to an appropriate volume (11 ?l) to accommodate in the total volume of hybridization mixture (36 ?l), as recommended by the array manufacturer (W.M. Keck Foundation Biotechnology Resource Lab at Yale University, New Haven, CT, USA). The prepared cDNA targets were kept in each tube separately on ice until mixed in the hybridization mixture.
2.5 Hybridization of Labeled cDNA Targets on Array Probes
2.5.1 Description of cDNA Microarray
Two Arabidopsis thaliana cDNA microarray slides (AR12K, serial #980 and #981; W.M. Keck Foundation Biotechnology Resource Lab at Yale University, New Haven, CT, USA) were used for the cross-species hybridization. Each array contained 11,960 expressed sequence tags (ESTs) generated from lambda PRL-2 cDNA library, which was cloned in the Plant Research Laboratory (PRL) at Michigan State University, East Lansing, MI, and made available from the Arabidopsis Biological Resource Center (ABRC) at The Ohio State University (http://www.biosci.ohio-state.edu/~plantbio/Facilities/abrc/abrchome.htm). The clones were prepared from the mixture of four types of tissues of the Columbia wild type of A. thaliana (L.) Heynh.: 1) etiolated seedlings; 2) roots; 3) rosette plants of various ages; 4) stems, flowers, and siliques at all stages from floral initiation to mature seeds (Newman et al., 1994). The cDNA probes were printed in 32-pin conformations on 1 inch x 3 inch glass slides. Each array contained 32 subarrays in a 4 x 8 block format, and each block had 16 rows and 24 columns.
2.5.2 Denaturation and Prehybridization of Arrays
Since the printed cDNA probes were double-stranded and the array slides were delivered without denaturation during post-print processing, the cDNA probes on the array slides were denatured to create single-stranded cDNA probes before hybridization with the labeled targets (Schena, 2003). The denaturation of cDNA probes was immediately followed by prehybridization. The purpose of prehybridization was to coat the surface of the glass to block any sites on which the fluorescently labeled target cDNA might bind nonspecifically and produce background signal (Anderson, 1995). The prehybridization solution contained a blocking agent, a detergent and random-sheared foreign DNA, which is heterologous to both labeled cDNA targets and printed probes on the array slide so as to minimize nonspecific binding of labeled cDNA targets to the slide (Table 2-5). Denaturation and prehybridization were carried out according to the microarray manufacturer's recommended protocol with some modification to adapt the availability of equipment. Briefly, 36 ?l of prehybridization solution was placed on each array slide, covered with a clean standard glass coverslip (22 x 50 mm), and laid on the In situ Adapter (Cat. #950007052; Eppendorf N.A., Westbury, NY), which in turn was fit onto the metal block of a thermocycler (Mastercycler? Gradient; Eppendorf N.A., Westbury, NY). Before placing the slide on the adapter, a thin layer of water was spread, to ensure even heat transfer. The thermocycler was programmed to heat the block to 76?C for 2 minutes and then to hold the temperature at 50?C.
Immediately after the temperature dropped to 50?C, the slides were transferred into aluminum hybridization chambers (Cat. #DT-1001; Die-Tech, San Jose, CA; Figure 2-4) and 300 ?l of pre-warmed chamber buffer (Table 2-6) was added in the grooves and reservoirs of each hybridization chamber to prevent evaporation of prehybridization solution from the slides. After tightening the thumb screws evenly, the hybridization chambers were placed horizontally in a 50?C water bath. After the array slides were incubated for 1 hour, the slides were carefully and quickly taken out of the hybridization chamber and immediately transferred to a glass Couplin jar (Cat. # 900570; Wheaton Sci. Prod., Millville, NJ) containing distilled-deionized water.
After the coverslips floated off the slide into the water, the array slides were carefully lifted avoiding contacts with the coverslips so as not to scratch the array, and were transferred to another Couplin jar containing fresh water. Prehybridization buffer was removed by gently agitating in the water for 2 minutes. The slides were then dehydrated by placing in 70% ethanol followed by 100% for 2 minutes each. The residual alcohol was evaporated in the air and the slides were kept in a vacuum-sealed canister (FOODSAVER? Round Canister, Cat. #T16-0032; Jarden Corp., Rye, NY) containing dust-free desiccants (DriCan? Reusable Desiccating Canister, Cat. #19950; Ted Pella, Inc., Redding, CA) while hybridization mixtures were being prepared. According to the array manufacturer, removal of the prehybridization buffer results in a more uniform and reproducible hybridization (W.M. Keck Foundation Biotechnology Resource Lab at Yale University, 2006). Denaturation and prehybridization of cDNA probes on the slide were carried out while preparing the hybridization mixture (described in the next section, "2.5.3 Preparation of Hybridization Mixtures" ), so that the labeled target could be applied immediately as soon as it was ready.
2.5.3 Preparation of Hybridization Mixtures
To compensate dye-related bias, which is commonly observed in the two-color microarray platform, dye-swap experimental design was used. In this experiment, one array was co-hybridized with shoot apex and young leaf cDNA targets that were labeled with Alexa Fluor? 555 (fluoresces green) and Alexa Fluor? 647 (fluoresces red), respectively. In a second co-hybridization, the dyes of the two samples were switched.
To prepare the two sets of hybridization mixture, 22.2 ?l hybridization buffer (Table 2-7) and 1.4 ?l blocking solution (Table 2-8) were added in each of two separate 0.5 ml microcentrifuge tubes. The blocking solution was intended to inactivate reactive groups remaining on glass microarray slide surface. It reduces background noise while maintaining full signal intensities for DNA microarray applications according to the microarray manufacturer (W.M. Keck Foundation Biotechnology Resource Lab at Yale University).
Into one of the two microcentrifuge tubes containing hybridization buffer and blocking solution, the shoot apex cDNA labeled with Alexa Fluor? 555 and young leaf cDNA labeled with Alexa Fluor? 647 were added. Into the other, shoot-apex and young leaf cDNA with reversed fluorophores were added (Table 2-9). The total volume of the hybridization mixture was 36 ?l.
2.5.4 Denaturation of cDNA Targets & Hybridization
The fluorescently labeled cDNA targets in the hybridization mixture were denatured at 90?C for 3 minutes on a thermocycler (Mastercycler? Gradient; Eppendorf N.A., Westbury, NY) and kept at 42?C for a short period of time until applied on the array slides. After mixing well by repeated pipetting, each of the hybridization mixtures was applied on each cDNA array slide with dye-swap (Table 2-10; Figure 2-5) carefully so as not to create bubbles, and covered with clean standard 22 x 50 mm glass cover slips. (The array was 18 x 36 mm in size and could be covered under 22 x 40 mm coverslip; but, it was not large enough to hold 36 ?l mixture.) After transferring the slides into the aluminum hybridization chambers (Cat. #DT-1001; Die-Tech, San Jose, CA), chamber buffer was added, as previously described in the section, "2.5.2 Denaturation and Prehybridization of Arrays" on page 40. The hybridization chambers were sealed by tightening the thumb screws, placed horizontally and incubated in a water bath at 42?C for 9 hours for hybridization.
2.6 Washing and Drying Hybridized Microarray
While hybridization was in progress, a series of wash solutions containing saline-sodium citrate (SSC) and sodium dodecyl sulfate (SDS) in decreasing concentrations (Table 2-11) was prepared in glass staining dishes (Slide Staining Dish with Removable Rack, Cat. #900200; Wheaton Sci. Prod., Millville, NJ) and pre-warmed at 32?C in a water bath, which was 10?C below the hybridization temperature.
After hybridization, while maintaining the level, the hybridization chambers were transferred onto a slide warmer, which was set to the same temperature as for hybridization. After the moisture on the outside of the hybridization chamber was wiped dry (especially in the gap between the upper and lower pieces of the chamber that were created by the thickness of the O-ring seal), the thumb screws were unfastened and the cover of the hybridization chamber was carefully prised open using a blade of the coverslip forceps. With the slide still in the hybridization chamber, the cover slip was removed with a fine forceps and the slide was placed as quickly as possible in the first wash dish containing 2X SSC and 0.1% SDS. Slides were washed for 10 minutes with gentle shaking, and transferred to the second wash dish containing 0.2X SSC and 0.1% SDS. After washing in the same way as in the first wash, the slides were transferred to the third wash dish containing 0.2X SSC, but no SDS, and washed for 10 minutes with gentle shaking. Finally, the last wash step was repeated with a fresh solution in another wash dish to ensure all residual SDS was removed.
To dry the slides, each slide was separately placed with printed side down in 50 ml BD Falcon? polystyrene conical tubes (Cat. #352073, BD Biosciences, San Jose, CA) separately and spun at 1000 rpm for 5 minutes on a bench-top centrifuge with a swing-bucket rotor (Centrific? Model 225; Fisher Scientific, Springfield, NJ). Dry slides were placed in a vacuum-sealed canister (FOODSAVER? Round Canister, Cat. #T16-0032; Jarden Corp., Rye, NY) containing dust-free desiccants (DriCan? Reusable Desiccating Canister, Cat. #19950; Ted Pella, Inc., Redding, CA) to transport to Laboratory of Genomics Bioinformatics & Proteomics, University of Toledo Health Science Campus, where scans were performed (described below).
2.7 Scanning Array Slides and Image Analysis
2.7.1 Scanning for Image Acquisition
Prepared array slides were scanned using a two-channel confocal microarray scanner (ScanArray?) and quantitated by the scanner's dedicated software (ProScanArray? Express v.3.0.0; ProSAE), which were bundled in a PerkinElmer? Microarray Analysis System (PerkinElmer Life and Analytical Sciences, Shelton, CT). To achieve optimal fluorescence intensity, the photomultiplier tube (PMT) gain was set at 80% of the maximum for both Alexa Fluor? 555 and Alexa Fluor? 647. The laser power for scanning was limited to 80% for Alexa Fluor? 555 and 84% for Alexa Fluor? 647, respectively, to reflect the difference in fluorescing capacity of the two dyes, while minimizing the photo-bleaching effects (PerkinElmer Life Sciences Inc., 2002). After laser focusing and balancing of the two channels, scans were conducted at a resolution of at 10 ?m. For each array scan, two separate 16-bit Tagged Image File Format (TIFF) images were produced and combined together to produce a composite image.
2.7.2 Array Spot Recognition ('Gridding')
For spot recognition, the grid was defined using the GAL (GenePix? Array List) file, which was provided by the array manufacturer (file name: AR12K-768+.gal; W.M. Keck Foundation Biotechnology Resource Lab at Yale University, New Haven, CT). The format of the GAL file was originally implemented by Molecular Devices (Union City, CA) and the file describes the dimension and position of blocks, the layout of spots, and the names and identifiers of the printed cDNA associated with each spot (Molecular Devices Inc., 2001; Zhai, 2001). During the spot finding procedure, the ProSAE software recognizes spots in 5 different classes of the spots based on the quality of the spots, ranging from 1 to 5. Flag code 1 indicates a spot that was 'not found'; flag code 2, 'found'; flag code 3, 'good'; flag code 4, 'bad'; and flag code 5, 'absent', which is not a spot on the array format. The flag code 2 was not seen because if the spot was found, it was either good or bad spot. Based on the flags, the 'gridding' process was repeated until the maximum number of spots on the array matched with the grid and recognized as good spots. As a final step of the spot finding procedure, the spots with artifacts, such as foreign particles, were marked as 'bad' and they were excluded from data analysis.
2.7.3 Array Spot Segmentation
After grids were properly placed, segmentation was carried out, in which foreground (spots) and background on the scanned images were defined and the pixel intensity data within the array spots were extracted. The 'adaptive circle segmentation' method was chosen in the ProSAE software. In this segmentation method, the program tries to find the edges of a spot and draws a circle around the spot, and what is inside the circle is the foreground (spots), and areas outside of the circle are background (Weeraratna and Taub, 2007). This method allows for the radius to be adapted to the spot shape and more accurate segmentation, compared to the 'fixed circle segmentation', in which spots are assumed to be circular with fixed radii (Li et al., 2005; Nagarajan, 2003; Rueda and Qin, 2004; Rueda and Qin, 2005; Wu et al., 2005).
2.7.4 Quantitation of Array Spot
After segmentation, the color information in the array spots were quantitated into intensity data. Background-corrected spot intensities were obtained by using a 'local background correction' method, where the mean intensity of pixels in the background was subtracted from the mean intensity of those in the spot.
2.8 Data Analysis
2.8.1 Selection of Data
Using the background-corrected mean spot intensity data, which were generated by ProSAE software, the patterns of differential gene expression in shoot apices and young leaves were analyzed. For analyses, data from spots that were marked as 'bad' were excluded from further analysis. Among 12,288 spots on each array, 336 were marked as "BLANK" in the GAL file supplied by the array manufacturer, and they were also excluded. "BLANK" means that AIG Locus link or gene description have not been found for the corresponding GenBank Accession.
When background-subtracted spot intensities became negative, those were regarded as missing values and excluded from the data analysis (Hovatta et al., 2005). If a spot had negative spot intensity in a channel, all other corresponding spots across the channels and arrays also were excluded from the data analysis.
2.8.2 Log Transformation and Normalization of Data
The spot intensity data were imported into GeneSpring GX software (Version 10; Agilent Technologies, Inc., Santa Clara CA, USA) and the data were transformed to their log2 values followed by Quantile normalization. A log2 transformation converts the expression values into an intuitive linear scale that represents two-fold differences (Alba et al., 2004). Quantile normalization makes the distribution of expression values of both channels and all samples in an experiment the same (Bolstad et al., 2003; Yang and Thorne, 2003). Thus, after the quantile normalization, all statistical parameters (i.e., mean, median and percentiles) of the sample become identical. Quantile normalization reduces variance between arrays, thus overcoming the differences among the arrays that are caused by non-biological factors, including dye-bias (Ewens and Grant, 2005; Hovatta et al., 2005). Quantile normalization was performed by the following steps (Agilent Technologies Inc., 2008; Mayer and Glasbey, 2005):
1. The spot intensity values of each sample were sorted in ascending order and placed next to each other.
2. Each column was sorted in ascending order. The mean of the sorted order across all samples was taken so that each row in the sorted matrix had value equal to the previous mean.
3. The modified matrix, as obtained in the previous step, was rearranged to have the same ordering as the input matrix.
2.8.3 Selection of Differentially Expressed Genes
The final step in the data analysis was to identify the genes that were differentially expressed in the shoot apices and the young leaves. When the log2 ratio in the spot intensity level between two tissue types was greater than 1 (i.e., 2-fold difference or larger), the corresponding gene was considered differentially expressed. A Student's t-test (p=0.05) was used to select the genes that were expressed differentially in the two tissue types with a statistical significance (Glantz, 2005). To facilitate finding the differentially expressed genes, a 'volcano plot' was produced in GeneSpring GX software, in which the fold-difference of expression level and corresponding p-value were plotted for easy identification of genes that fall into the selection criteria, which were log2 ratio >1 and p=0.05.
2.9 Functional Analysis of Differentially Expressed Genes
2.9.1 Cluster Analysis
To identify and group together the genes that were similarly expressed and infer any biological significance of the group of genes, cluster analyses were carried out using two methods provided by GeneSpring GX: hierarchical clustering and K-mean clustering. In hierarchical clustering, the co-regulated genes were grouped by distance matrix calculated based on the Euclidian distance. Hierarchical clustering does not distribute data into a fixed number of clusters, but produce a grouping hierarchy so that most similar entities are merged together to form a cluster.
In contrast to hierarchical clustering, in K-mean clustering, genes are partitioned into a fixed number (k) of clusters such that, genes within a cluster are similar, while those across clusters are dissimilar. Based on the number of clusters obtained from the hierarchical clustering analyses, K-mean clustering was also carried out to compare the outcomes between the two clustering methods. To begin with K-mean clustering, genes were randomly assigned to four distinct clusters and average expression vector was computed for each cluster based on the Euclidean distance as in hierarchical clustering analyses. For every gene, the algorithm then computed the distance to all expression vectors, and moved the gene to that cluster whose expression vector was closest to it. The entire process was repeated iteratively until no genes can be reassigned to a different cluster, or 50 iterations were reached.
2.9.2 Functional Annotation
Differentially expressed genes in either tissue type were categorized based on the functional annotations. Since GenBank accession numbers for ESTs (Expressed Sequence Tags) were used to identify the array features in the GAL file provided by the array manufacturer, they were first converted (mapped) into AGI (Arabidopsis Genome Initiative) locus identifiers using an 'association' file downloaded from the TAIR ftp site (ftp://ftp.arabidopsis.org/home/tair/Genes/TAIR9_genome_release/). These locus identifiers were used for subsequent query and retrieval of the gene descriptions from The Arabidopsis Information Resource (TAIR; Berardini et al., 2004; http://www.arabidopsis.org/tools/bulk/go/index.jsp).
Using the GO Slim Classification for Plants at TAIR, the genes that were statistically differentially expressed in the shoot apex or the young leaf were functionally categorized. The TAIR GO slim is a reduced set of GO terms from the Gene Ontology (GO; The Gene Ontology Consortium, 2000; http://www. geneontology.org/), which was tailored to Arabidopsis plants and useful to provide a broad view of a given gene set. To obtain the low-level GO categories for the detailed annotation, a subset of the TAIR Genome (Release 9) with matching locus identifiers was downloaded. Since the high- or low-level GO categories resulted in too broad or too narrow of classifications in the functional annotations, the intermediate categories were retrieved manually using the individual link to each GO terms in the TAIR genome database.
The genes were categorized manually and described based on the three GO vocabularies, each providing a specific type of information about the gene or protein: (i) the pertinent biological processes, (ii) its specific molecular function, and (iii) its cellular localization. To fine-tune the categorization the gene products were also queried in the protein databases, such as the Universal Protein Knowledgebase (UniProtKB; The UniProt Consortium, 2009; http://www.uniprot.org/) and the Munich Information Center for Protein Sequences (MIPS; Mewes et al., 2008; http://mips.helmholtz-muenchen.de/). The functional role of the uncharacterized gene products in these databases were predicted based on their occurrences of their functional domains provided in the InterPro database (Hunter et al., 2009; http://www.ebi.ac.uk/interpro/) and supporting publications.
2.9.3 Pathway Analysis
To see if there was any unifying biological theme in the gene set, which was obtained from the statistical analysis, pathway analysis was carried out in GeneSpring GX using the parameters set as in Table 2-12.
2.9.4 Verification of Differential Expression of Genes
Instead of using RT real-time PCR (RT-rt-PCR) to verify the expression levels of the genes that were predominantly expressed in a tissue type, a microarray data exploring tool, the electronic Northern and the electronic Fluorescent Pictograph (eFP) Browser (http:// www.bar.utoronto.ca/) were used to explore the expression levels with data set obtained from other experiments (Winter et al., 2007). This eFP Browser engine paints data from large-scale data sets onto pictographic representations of the experimental expression data from the AtGenExpress Consortium to explore the expression levels among different plant parts and the developmental stages (Goda et al., 2008; Kilian et al., 2007; Schmid et al., 2005).