Caltechdeki staj sonunda sundugum rapor
SURF Final Report
Identification of Transcription Factors That Recognize the A-box, a DNA Sequence Required for Dorsolateral Patterning in Drosophila
Fatih Kocabaş
Mentor: Angelike Stathopoulo (co-mentor Phoebe Tzou)
Patterning in the early Drosophila embryo is regulated by transcription factors. These transcription factors act in a sequence specific manner by binding to sites within enhancers or silencers, control regions of genomic DNA that regulate the expression of genes. We have focused on one such motif, ATTCATTCATGA, a conserved sequence which we call the A-box. Our aim was to identify the transcription factor(s) which bind this sequence. The A-box is a novel dorsolateral repression element found within an enhancer for a gene expressed along the dorsoventral axis, the intermediate neuroblast defective (ind) gene. It is necessary to mediate repression of ind expression within dorsal regions. We used a one-hybrid approach to isolate the DNA-binding proteins that recognize this site. At the beginning of this study, cDNA from 0-4 hour’s early (wild type) embryo of Drosophila was prepared and it was tagged with a transcriptional activation domain by insertion into a specific vector (pGADT7-Rec2). By constructing a library from properly staged embryos we hoped to eliminate background and maximize the probability of obtaining true positives. This library was then transformed into yeast strain together with an A-box reporter vector allowing selection based on Histidine auxotrophy (pHIS2 vector with the A-box sequence driving HIS3 expression). We also added an additional screen by creating a second reporter based on lacZ expression. We used the pJL638 vector with A-box sites inserted upstream of lacZ; this vector was integrated into wild-type yeast strain (W303). We screened for library plasmids that bind to the reporter and activate HIS3 as well as lacZ expression. We isolated plasmids from library transformants that passed these two conditions, in order to identify the cDNA insert. We expected to find 1 or 2 transcription factors that bind ATTCATTCATGA with high affinity. This work was important because no other repressor functioning in dorsal regions of the Drosophila early embryo has been characterized to date.
Drosophila development results from a transcriptionally controlled cascade. The sequence is largely controlled at the levels of transcription and translation. Although early embryos do not carry out transcription, after several cell divisions, mRNA production begins, form the mRNA’s needed for later development. Some of them provide positional information, extend of gene expressions and define border of body segments. At the activity of these mRNAs and transcription factors, enhancers take important roles in pattern formation. In the one of the recent studies in the Stathopoulos laboratory, it was found that the ind enhancer, which affects dorsal and lateral patterning in Drosophila, has a novel silencer element, and it also contains various activator and repressor elements1. It is necessary to mediate repression of Dorsal-target gene expression in the dorsal ectoderm of Drosophila. We already know that while ventral limits of gene expression are defined by the Snail and Vnd repressor, dorsal border is defined by A-box. However, we do not know what the transcription factors are binding this site. We should know that to make clear dorsaletaral gene network.
In this study, we studied the silencer element – ATTCATTCATGA- called A-box found within the module B of ind enhancer and tried to identify and isolate transcription factor(s) bind to this conserved element by using one-hybrid assays. We used One-Hybrid assays because it enables us to study transcription factor(s) that bind(s) to a target element that activates transcription from minimal promoter as shown in figure 1. In this method, DNA-binding proteins like transcription factors are expressed as hybrid proteins, in other words fusions to GAL4 activation domain (AD) by cloning the cDNA into pGADT7-Rec2 vector (Figure 2). This vector is a low-copy number vector, it is Sma1 linearized and it has special ends to mediate homologues recombination with cDNA in yeast cells by using cells’ recombinases. Binding of hybrid protein to the target sequence activates transcription of HIS3, makes possible to grow yeast strain Y187, a His auxotroph, on minimal medium lacking histidine.
To make experiments rapid, we used cotransformation method to construct AD fusion library. Moreover, to find the optimal 3-AT concentration to suppress leaky HIS3 expression in reporter plasmid pHIS2 with target sequence, we also used control vectors such as p53HIS2 and pGAD-Rec2-53 (Figure 4-5) and cotransform with reporter pHIS2 vector and plate on a more stringent medium SD/-His/-Leu/-Trp than SD/-His/-Trp.
Because it is always possible to see leaky expression of HIS3 gene, we tried to find the optimum 3-AT concentration. We found that 5 mM 3-AT concentration was enough to suppress leaky histidine production on SD/-His/-Leu/-Trp agar plates. However, to make conditions more stringent and to look at stronger interactions with our sequence, we chose 7 mM because cotransformed yeast cells with p53HIS2 Control vector (figure3) and pGAD-Rec2-53 control vector (figure4) could easily form colonies at 10 mM 3-AT concentration. But, p53HIS2 Control vector contains 3 copy of p53 DNA element so there was more activation than one DNA element. We cloned into pHIS2 only one DNA element, therefore 7 mM were affective as much as 20 mM, so we could see high affinity bindings to our target element ATTCATTCATGA.
On the other hand, we also performed LacZ expression experiments to see LacZ induction by the corresponding library binding protein2. Therefore, we designed new oligonucleoties to clone into pJL368 vector containing LacZ. In this system, we activated transcription of LacZ instead of HIS3. Moreover, pJL368 vector contained Ura2 gene, so it enabled us to select transformants on minimal media lacking uracil amino acid by using a new yeast strain which was ura-, so it could not grow on media lacking essential amino acid uracil. By looking color of colonies formed, as described in figure 3, we decided which colony contain cDNA insert expressing binding transcription protein. Similar with previous study, an expression library of hybrid proteins were formed by using Sma1 linearized pGADT7-Rec2 vector and cDNA. These hybrid proteins recognized the binding site and acted as transcriptional activators of the reporter gene and they turned cells blue in a β-Galactosidase assay. Moreover, we also selected them by using 3-AT suppression.
Finally, we confirmed positive colonies by performing β-Galactosidase assay on colonies that grow on medium that lacks histidine due to reporter activation. We studied on 35 positives and isolated plasmids. After that we transformed them into bacteria to select our plasmid containing cDNA insert. Up to now, we could not get any valuable result. Although, there are some inserts bigger than 1.8 kb, most of them are around 1.0 kb which we consider them as false positives, not real one-hybrid candidates. However, we expect to find 1 or 2 transcription factors that bind A-box with high affinity, therefore I suggest continuing experiments and one-hybrid screens by considering possible errors or critical points of study. First of all, total number of transformants is around 19500. This is not an enough number for a library search, I expected to see more at least 50000 transformants per cDNA prepared. One of the reasons might be that mRNAs expressing transcription factors related our sequence might have short half-life, so they degraded rapidly because library efficiency was not as expected. If it is so, then, while production of cDNA, to get more reliable cDNA, we should use more stringent methods to save mRNA. By using this new cDNA produced from this total mRNA, we should repeat one-hybrid screens. Moreover, it is also possible that His selection is not stringent enough. To increase stringency of His selection we can use plates containing up to 25mM 3-AT concentrations instead of 7mM. Last but not least, it might be necessary to make collection of embryos evenly. There should be embryos from every stage from 0 hour to 4 hours.
Results of this work would provide information about first repressor functioning in dorsal regions of the Drosophila early embryo which has not been characterized to date. Besides, if we could get any transcription factors, we could continue other experiments like hybridizations related with these transcriptions factors, and we could further clarify dorsaletaral gene network in early Drosophila embryo. Finally, use of second reporter like LacZ reporter was very helpful to be sure about positives, because we used only strongest color signal produced after B-gal assay.
METHODS
Yeast One-Hybrid Screens. We performed yeast one hybrid screen to identify transcription factors. The basic protocol used for yeast one-hybrid screening has been described in BD Matchmaker™ Library Construction & Screening Kits User Manual3. Because of the large number of false positives obtained with this procedure, we designed another reporter use in B-Galactosidase assay to more easily and accurately identify the true positives. The HIS3 reporter plasmids, pFK201, pFK203, and pFK205, were constructed by inserting one copy of Oligo A, Oligo B and Mutant Oligo B, respectively, into pHIS2 cloning vector which is linked to a minimal promoter of HIS3 locus (PminHIS3). These plasmids were used in histidine selection. On the other hand, the LacZ reporter plasmids, pFK1, pFK2, and pFK3, were constructed by inserting one copy of Oligo A’, Oligo B’ and Mutant Oligo B’, respectively, into pJL638. These plasmids were integrated into the genome of yeast strain W303, forming the FK1, FK2, and FK3 yeast strains, respectively. cDNA is synthesized by using BD SMARTTM technology described in BD Matchmaker™ Library Construction & Screening Kit. Then, recombination-mediated cloning is used to fuse BD SMART ds cDNA with the GAL4 AD. Three dual reporter yeast strains were then constructed by cotransformation. By using cotransformation protocol, GAL4 AD Fusion Libraries is constructed by intracellular recombinase system of yeast into the pGADT7-Rec2 vector and pFK201, pFK203, and pFK205 vectors were transformed into the FK1, FK2, and FK3 yeast strains, respectively. Total 19500 transformants were grown on SD/-LEU/-HIS /- TRP+7mM 3-AT plates and subjected to an X-Gal screen. Clones encoding transcription factors were identified using 3-AT and X-Gal selection. Survival of transformants in the absence of leucine and trptophan is due to the LEU2 gene present in library vector, pGADT7-Rec2, and TRP1 gene present in pHIS2. Cotrasnformants (pFK201, pFK203 into FK1, FK2 yeast strains respectively) containing library vectors that encode for a true binding factor turned blue after X-Gal Assay and gave total 35 true blue colored clones. Moreover, FK3 yeast strain remained white although transformed with pFK205 and GAL4 AD fusion library and cDNA.
Constructing a Reporter Vector:
We designed six antiparallel oligonucleotides, one representing the sense strand and the other its antisense complement so that we obtained OLIGO A, OLIGO B and Mutant OLIGO B. Firstly, we for each Oligo, we mixed 0.1 µg of sense-strand and 0.1 µg of antisense-strand oligonucleotide in 10 µl of 50 mM NaCl. Then, we incubated at 70oC for 5 min and then we slowly cooled to room temperature (~10 min). Moreover, we also designed Oligo A’, Oligo B’ and Mutant Oligo B’ to integrate into pJL638 vector and we followed same procedure for oligo synthesis.
Designed Oligonucleotides for pHIS2 vector
OLIGO A: PT201- PT202
5’GCAGCGCATTCATTCATGAGGCCAAGAGCT 3’
3’TCGACGTCGCGTAAGTAAGTACTCCGGTTC5’
OLIGO B: PT203-PT204
5’GGATGTATTCATTCATGAAGTGTCTGAGCT3’
3’TCGACCTACATAAGTAAGTACTTCACAGAC5’
And mutant OLIGO B: PT205-PT206
5’GGATGTATTAATTAATTAAGTGTCTGAGCT 3’
3’TCGACCTACATAATTAATTAATTCACAGAC5’
Designed Oligonucleotides for pJL638 vector
OLIGO A’:
5’ GATCGCAGCGCATTCATTCATGAGGCCAAG 3’
3’CGTCGCGTAAGTAAGTACTCCGGTTCCTAG 5’
OLIGO B’:
5’GATCGGATGTATTCATTCATGAAGTGTCT 3’
3’CCTACATAAGTAAGTACTTCACAGACTAG5’
And mutant OLIGO B’: A NEGATIVE CONTROL
5’ GATCGGATGTATTAATTAATTAAGTGTCT 3’
3’CCTACATAATTAATTAATTCACAGACTAG5’
To digest pHIS2 vector completely, we added 1 µl pHIS2 (500 ng/µl), 2 µl 10X NEB buffer #1, 2 µl 10X BSA, 1.5 µl Sac I enzyme and we completed up to 20 µl with 13.5 µl ddH2O. Then, we incubated at 37°C for 1.5 hr. Before passing ligation reaction, we electrophoresed a 2 µl of sample of the digest on a 1% Agarose gel at 150 volts to confirm that the plasmid has been efficiently linearized. In the ligation reaction, we used 5 µl of digested DNA for each construct, and we added 1 µl of annealed oligo (OLIGO A, OLIGO B and Mutant OLIGO B), 1.2 µl of 10X T4 ligation buffer (Roche), 0.8 µl (at least 0.8 units) of T4 DNA ligase (NEB) and 4 µl of ddH2O (total 12 µl). After completion of mixtures, we incubated it at room temperature for at least 4 hours. Afterwards, we separately transformed CaCl2 competent E. coli cells (DH5α strain) with each construct using a standard method. We transformed 6 µl of ligated plasmids by mixing 100 µl of CaCl2 competent E. coli cells thawed on ice. After incubation on ice for 15 minutes we made heat shock for 15 seconds at 42°C. Immediately, we removed mixture from heater and then we put on ice. On ice, we added LB medium up to 1ml (900 µl of LB). Following step was shaking at 165rpm for 1 hour to recover cells. We also transformed 0.5 µl of pHIS2 SacI digest as a negative control. After 1 hour, we plated transformants on LB/kan plates, and incubated at 37°C overnight.
Integration of Oligo A’, Oligo B’ and Mutant Oligo B’ into pJL638 vector and integration of these new vectors into genome of W303:
After oligo synthesis, we ligated them into Bgl II site of Bgl II digested plasmid. After obtaining transformants into E. coli and growing on LB/amp plates, we checked isolated plasmids via Zyppy Plasmid Miniprep Kit by double digestion with Stu I and Bgl II restriction enzyme. We expected to a digest about 500 bp.
After checking plasmids whether they contain our target sequence by sequencing, we digested the plasmid from Stu I site, which is found in the Ura gene, and using 1 µg pJL368-Stu I digest we integrated plasmid into genome of W303 yeast strain. We used following formula for integration:
Herring testis Carrier DNA 10 µl
pJL368-Stu I digest 1 µg
W303 competent cells 300 µl
PEG/LiAc 1000 µl
--------------------------------------
Then we added 80 µl of DMSO.
After integration into genome of W303, we selected positives on SD/–ura plates.
Screening Colonies by PCR:
We selected 8 colonies from transformants for each Oligo. We isolated plasmid DNA by using Zyppy Plasmid Miniprep Kit. Then we performed PCR mixture and reaction by using following conditions:
Miniprep DNA 1 µl
10X Taq buffer 5 µl
MgCl2 2 µl
dNTPs [2.5mM] 2 µl
Forward Primers: PT201, PT203, PT205 [10pM] 1 µl
Reverse Primer: PT214 [10pM] 1 µl
Taq Pol. 0.5 µl
ddH2O 37.5 µl
----------------------------
Total 50 µl
PCR reaction: Put Oligo A to 65.9°C, Oligo B to 57.9°C, and Mutant Oligo B to 54.4°C wells.
Step 1: 94°C 1 min
Step 2: 94°C 30 sec
57-68°C 50 sec
72°C 40 sec
Step 3: Return Step 2 for 34 times
Step 4: 72°C 7 min
4°C forever
After completion of PCR, we electrophoresed 8 µl of PCR product on 1.2% Agarose Gel to see and selected positive colonies depending on the expected PCR product. Afterwards, we sent 0.2 µg/µl plasmids of selected colonies for sequencing and then we checked any mutation and direction of insertion.
RNA SOURCE. We collected early Drosophila embryos (wild type) which were 0-4 hour’s embryos. After obtaining enough amount of embryo, we isolated total RNA. Then, we used this RNA to prepare cDNA.
Collection of Drosophila Embryos:
We used 0-4 hour Drosophila embryos. We washed embryos by 1X Triton/NaCl solution by brush, and then bleached 3 minutes on plated. We again washed by 1X Triton/ NaCl solution. To remove Cl we washed under water, then we opened collector and transferred into a 1.5 ml of centrifuge tube containing 1.0 ml of ddH2O. We centrifuged at 13.2 rpm for 30 seconds. (Remove water)
Total RNA isolation from Drosophila Embryo:
We used procedure supplied by Dr. Phoebe Tzou. First of all, we added µl of 200 TRIzol and homogenized embryos completely. Again we added 200 µl of TRIzol. After incubation at RT for 5 minutes, we centrifuged at 13.2rpm at RT. Then we took supernatant into clean tube (~400 µl), added 100 (24:1) isoamyl alcohol, and mixed vigorously. We spinned mix at 4 °C, 13.2 rpm for 15 minutes. Without touching interface, we transferred aqueous phase to a clean tube and added 0.7 volume of isopropanol. After that, we incubated at RT for 10 minutes. To remove supernatant, we spinned at 13.2 rpm for 10 minutes at 4°C. Following removal of supernatant, we washed with 75% RNase-free Ethanol (~500 µl. We vortexed and centrifuged at 13.2 rpm, RT. After removal of ethanol, we air dried it for 2-3 minutes, then we added RNase-free deionized water. We resolubilized pellet by gently by vortex and pipette up&down to ascertain that pellet is resolubilized completely. We also prepared RNA gel to see result of total RNA isolation. Finally, we stored total RNA at -80°C to use at cDNA preparation.
We synthesized First-Strand cDNA using a Random Primer provided BD Matchmaker™ Library Construction & Screening Kit3 by as following:
1. Combine the following reagents in a sterile 0.25-ml microcentrifuge tube:
2 μl RNA sample, (For the control reaction, use 1 μl [1 μg] of the control RNA.)
1.0 μl CDS III/6 Primer
2 μl Deionized H2O
---------------------
4.0 μl Total volume
2. Mix contents and spin briefly.
3. Incubate at 72°C for 2 min.
4. Cool on ice for 2 min.
5. Spin briefly.
6. Keep the tube at room temperature and add the following:
2.0 μl 5X First-Strand Buffer
1.0 μl DTT (20 mM)
1.0 μl dNTP Mix (10 mM )
1.0 μl MMLV Reverse Transcriptase
9.0 μl Total volume
7. Mix gently by tapping. Spin briefly.
8. Incubate for 10 min at room temperature.
9. Incubate at 42°C for 10 min.
10. Add 1.0 μl BD SMART III Oligonucleotide.
11. Incubate at 42°C for 1 hr in an air incubator or hot-lid thermal cycler.
12. Place the tube at 75°C for 10 min to terminate first-strand synthesis.
13. Cool the tube to room temperature, then add 1.0 μl (2 units) RNase H.
14. Incubate at 37°C for 20 min.
15. If you plan to proceed directly to the LD-PCR step, take a 2-μl aliquot from the first-strand synthesis and place it in a clean, prechilled, 0.5-ml tube or place at –20°C. (First strand cDNA can be stored at –20°C for up to three months.)
We also amplified ds cDNA by Long Distance PCR (LD-PCR) again as described in BD Matchmaker™ Library Construction & Screening Kits Manual3:
1. Preheat the PCR thermal cycler to 95°C.
2. To prepare sufficient ds cDNA for transformation, set up two 100-μl PCR reactions for each experimental sample. Set up one reaction for the Control sample. In each reaction tube, combine the following components:
2 μl First-Strand cDNA
70 μl Deionized H2O
10 μl 10X BD Advantage 2 PCR Buffer
2 μl 50X dNTP Mix
2 μl 5' PCR Primer
2 μl 3' PCR Primer
10 μl 10X GC-Melt Solution
2 μl 50X BD Advantage 2 Polymerase Mix
-----------------------
100 μl Total volume
3. Mix gently by flicking the tube. Centrifuge briefly.
4. Overlay the reaction mixture with 2 drops of mineral oil if necessary. Cap the tube and place it in a preheated (95°C) thermal cycler.
5. Begin thermal cycling. If you have a hot-lid thermal cycler, use the following program:
• 95°C 30 sec
• 21 cycles:
95°C 10 sec
68°C 6 min+5sec
• 68°C 5 min
6. When the cycling is complete, analyze a 7-μl aliquot of the PCR product from each sample alongside 0.25 μg of a 1-kb DNA size marker on a 1.2% agarose/EtBr gel.
7. Purify ds cDNA by using BD CHROMA SPIN™ Column and Collection tubes or store ds cDNA at –20°C until use.
Cotransform Yeast Strain Y187 with ds cDNA, pGADT7-Rec2, and pHIS2/target DNA. (Same protocols are used as given in the BD Matchmaker™ Library Construction & Screening Kits Manual)3
1. Prepare competent yeast cells (large scale).
2. In a sterile, 15-ml tube combine the following:
• 20 μl ds cDNA (from Section IX.I, Step 16)
• 6 μl pGADT7-Rec2 (0.5 μg/μl)
• 5 μg pHIS2 target DNA containing Oligo A, B and mutant B.
• 20 μl Herring Testes Carrier DNA, denatured*
* Transfer ~50 μl of Herring DNA to a microcentrifuge tube and heat at 100°C for 5 min. Then, immediately chill the DNA by placing the tube in an ice bath. Repeat once more before adding the DNA to the 15-ml reaction tube.
3. Add 600 μl of competent cells to the DNA.
4. Gently mix by vortexing.
5. Add 2.5 ml PEG/LiAc Solution.
6. Mix gently by vortexing for 3–5 sec.
7. Incubate at 30°C for 45 min. Mix cells every 15 min.
8. Add 160 μl DMSO, mix, and then place the tube in a 42°C water bath for 20 min. Mix cells every 10 min.
9. Centrifuge at 700 x g for 5 min.
10. Discard the supernatant and resuspend in 3 ml of YPD Plus Liquid Medium.
11. Incubate at 30°C with shaking at ~265 rpm for 90 min.
12. Centrifuge at 700 x g for 5 min.
13. Discard the supernatant and resuspend in 6 ml of NaCl Solution (0.9%).
At the same time, we also prepared One-Hybrid Controls by using p53HIS2 and p53HIS2 vectors as positive control and by using pGAD-Rec2-53 and pHIS2 vectors as negative control as described in the same manual.
Selection of One-Hybrid Interactions3
1. To determine the transformation efficiency and to calculate the number of clones screened, spread 100 μl of a 1:10, 1:100, and 1:1,000 dilution onto 100-mm SD/–Leu, SD/–Trp, and SD/–Leu/–Trp agar plates.
2. Spread the remaining mixture on SD/–His/–Leu/–Trp + optimal [3-AT] plates (150 μl cells/150-mm plate) to select for one-hybrid interactions.
3. Incubate at 30°C for 3–7 days until colonies appear.
4. Calculate the transformation efficiency and number of clones screened:
a. Colonies on SD/–Leu x dilution factor ÷ volume (ml) plated x 6 ml = #transformants per 3 μg pGADT7-Rec2. Expected: ≥1 x 106 transformants / 3 μg pGADT7-Rec2
b. Colonies on SD/–Leu/–Trp x dilution factor ÷ volume (ml) plated x 6 ml = #clones screened. Expected: ≥3 x 105 clones / library
5. Restreak the His+ colonies on fresh SD/–His/–Leu/–Trp + optimal [3-AT]. Incubate at 30°C for 2–4 days. Seal the plates with Parafilm and store at 4°C. For long-term storage, prepare glycerol stocks and store at –70°C.
B-Galactosidase assays
For LacZ expression analysis, we used Colony-Lift Filter assays (page 23) as given in the Clontech yeast protocols handbook. As a filter paper, we used Whatman #5 papers for both colony lifting and soaking by Z buffer/X-gal solution. I also used yeast plasmid isolation protocol given in the Clontech yeast protocols handbook to isolate yeast plasmid. Because library protein had ampr gene, we transformed yeast miniprep via Electroporation and plated them on LB/amp plates.
FIGURES
XXX
Figure 1: One-hybrid System, library protein from cDNA forms a hybrid protein with GAL4 Activation Domain protein. Then, this hybrid protein activates transcription of HIS3 so that cells with this reporter vector containing target sequence grow on minimal media lacking histidine whereas other with mutant target sequence do not.
XXX
Figure 2: pGADT7-Rec2 Vector, The BD Matchmaker One-Hybrid Library Construction & Screening Kit contains the Sma I-linearized form of this vector, the form used for recombination-mediated cloning in yeast.
XXX
Figure 3: Diagram for one-hybrid screen for identifying proteins that recognize a binding sequence of interest by using LacZ activation and β-Galactosidase assay. Reporter genes containing either wild-type (top, blue colony formation) or mutant (bottom) binding sites in their promoter regions were used to test the sequence specificity of the LacZ induction.
XXX
Figure 4: p53HIS2 Control vector
XXX
Figure 5: pGAD-Rec2-53 control vector
Acknowledgments:
The author thanks his Mentor Dr. Angelike Stathopoulos, supervisor Phoebe Tzou for discussions of project and great assistance in the experiments, and also to Desirea for supplying some essential items, and collection of embryo, finally, Zinn Laboratory for their allowance to use some of their laboratory equipments. He also very grateful Caltech and The Caltech Student-Faculty Programs Office for their sincere help and support my research by SURF award.
References:
1. Stathopoulos, A. & Levine, M. Localized repressors delineate the neurogenic ectoderm in the early Drosophila embryo. Dev. Bio. 280, 482–493 (2005).
2. Li, J.J. & Herskowitz, I. Isolation of ORC6, a component of the yeast origin recognition complex by a one-hybrid system. Science 262 (5141), 1870–4 (1993).
3. BD Matchmaker™ Library Construction & Screening Kits User Manual. Retrieved August 1, 2005 from http://www.clontech.com/clontech/techinfo/manuals/PDF/PT3529-1.pdf