Subcloning inserts from traditional BAC or PAC vectors into the BIBAC:
recommended protocol, CMH

Do a CsCl prep of the BIBAC vector.

Do a CsCl prep of the BAC clone of interest.

Cut the BAC clone with NotI. To my knowledge all of the BAC type vectors have NotI sites flanking the cloning site. [I have better luck using the entire BAC (vector + insert) digest rather than isolating the high molecular weight insert. My interpretation of this is that with fewer manipulations the insert is less damaged. This is only what works best for ME.]

Cut the BIBAC with NotI.

Quantify the BIBAC NotI DNA compared to the high molecular weight BAC NotI fragment. We use lambda DNA as a standard loaded at 20, 50, 100, and 200 ng on a standard gel.

Treat BIBAC NotI with SAP (shrimp alkaline phosphatase USB) or other phosphatase (CIAP BRL).

Set up a series of ligations using BIBAC vector with and without phosphatase and with a ration of vector: insert of about 10:1. I recommend that you set up a couple of different ratios because estimating the concentrations is not always accurate.

Here is an example of a (successful) ligation mix.

BIBAC2 vector NotI + CIAP 200 ng 4 microliters
BAC clone NotI (pBeloBAC or derivative)
= vector (CmR 7 kb) + insert (128 kb) 100 ng 7 microliters
10X ligation buffer (NEB) 10 microliters
T4 DNA ligase (NEB) 2 microliters
high quality water (distilled, deionized) to make 100 microliters 75 microliters

Mix DNA + vector + water (gently) incubate at 60 degrees C for 10 minutes then remove to room temp.

Mix in 10X ligation mix + enzyme incubate overnight at 16 degrees C.

Drop dialyze a portion of the ligation mix (30 microliter) against water at room temp for 30 minutes (not essential, but you will get more colonies from your electroporation).

Use 1 microliter for electroporation of DH10B cells.

When subcloning, I usually plate out on Kan and then replica plate to test for CmR (internal control for ligation) and Kan + sucrose and then print a Kan plate last.

You should get a few colonies that are CmR – they contain the BAC vector – but also serve to tell you that the ligation worked.

From this ligation mix I checked 38 colonies (from one electroporation); 10 were CmR, all were Kan + sucroseR. Apparently some CmR colonies that are sucrose sensitive have also been observed. It is possible that this is due to read through of the sacB gene when CmR gene is inserted upstream and in the same orientation – but this has not been verified.

You can also plate a second electroporation onto Kan + sucrose, but you should replica plate these as well. In some batches of competent cells we have observed contaminants that grow on Kan + sucrose and NOT on Kan. Also you can always manage to generate “garbages” (deletions, rearrangements) in any “bad” ligation mix.

Analyze CmS, Kan + SucR colonies by standard BAC mini-prep. For this experiment 16 were checked on a pulsed field gel. Eight of these that appeared to have an insert of the correct size were then cut with EcoRI and run on a standard gel to compare the restriction pattern with that of the original BAC clone.

Modifying the BIBAC vector: recommended protocol, CMH

(eg, introduction of a desired plant selectable marker into pCH20, or BIBAC1)

Do a 2 liter prep of the (BIBAC) plasmid to be modified (protocol included).

Restriction enzyme digestion is by standard methods.

Isolation (if needed) of fragment to be inserted into the BIBAC is by standard methods (eg prep-a-gene, gelase).

We use Klenow to “chew back” 3′ overhangs as well as to fill in 5′ overhangs to blunt ends as needed. We use standard reaction conditions for both activities.

Ligations are standard but are likely to be blunt-end ligations. For blunt end ligation s I recommend that PEG (10%) be included in the ligation mix or use a commercial T4 DNA ligase buffer that already has PEG (eg BRL) which serves as to exclude volume, increasing the concentration of ends in the reaction mix.

I usually set up a series of ligations.

I use 20-50 nanograms of vector in a 10 microliter (final volume) ligation mix. Fragment is added in a ratio of 5-10 molar excess of fragment (insert) to vector.

You can set these up with and/or without treating with phosphatase. In my hands CIAP is more “perfectly” efficient (and we use it for library construction), but for standard cloning I prefer SAP for the ease of heat inactivation and better activity on blunt ends.

[I usually set the ligations up both + and – SAP.]

You should include standard controls, in particular, vector no ligase.  Incompletely digested vector is a common source of background. In cloning experiments that have no selection for the desired product, this becomes significant.

When you are not using sacB as a selectable marker (or even if you are, see BAC subcloning section) – I recommend plating transformants on Kan plates and then replica plating to Kan + sucrose. This provides more accurate information (as there is always some background with any marker, and we have observed contaminants in competent cell preps that grow on Kan + sucrose that are not really Kan resistant). If cells with the BIBAC vector are plated on Kan + sucrose, you will get a few colonies that are likely to contain a mutation that prevents expression of sacB. This will occur at a low frequencies under strong selection pressure (survival).


“LB” = *Lennox L Broth Base (GIBCO BRL) – autoclave

formula per 1 liter demineralized water:

peptone 140 (pancreatic digest of casein) 10 g
yeast extract, autolyzed, low sodium 5 g
sodium chloride 5 g
For agar plates add:
bacteriological agar (2%) 20 g

* for E. coli (growth at 37 degrees C)
* for A. tumefaciens (growth at 30 degrees C)

SOC broth for recovery of cells following electroporation – autoclave formula per 1 liter demineralized water:

bactotryptone 20 g
yeast 5 g
NaCl 0.6 g
KCl 1.86 g
MgCl2 2.04 g
MgSO4 2.46 g
glucose 3.6 g

Following electroporation, the cells are pipetted from the electroporation cuvette into sterile tubes containing SOC media at room temperature.

Incubate E. coli @ 37 degrees C for ~ 40 minutes

Incubate A. tumefaciens @ 30 degrees C for ~ 60 minutes

High sucrose agar medium
For positive selection of inserts in BIBAC vectors (inactivation of sacB gene) add 100 mls of 50% sucrose (autoclaved or filter sterilized is ok) to 1 liter of Lennox L Broth based agar. Add required antibiotics taking into account that final volume is 1100 mls. 3-ketoglycoside test for Agrobacterium Based on a metabolic difference between the Agrobacterium genus and other bacteria such as Escherichia spp and Rhizobium spp. (Bernaerts and DeLey, 1963, Nature 197, 406-7) Agrobacterium spp grown on lactose make an enzyme called hexopyranoside cytochrome C oxidoreductase that can convert lactose into 3-ketolactose.

3-keto medium

for 1 liter, add to demineralized water: (final)
lactose 10 g 1% lactose
yeast extract 1 g 0.1% yeast extract
bacto-agar 20 g 2%

Streak strain on plates and incubate at 30 degrees C. After colonies appear, flood the plate with a shallow layer of Benedict reagent. Leave at room temperature. A yellow ring of Cu2O will appear wherever there is 3-ketolactose produced by Agrobacterium.

Note: in our experience the intensity of the yellow produced can vary widely between different Agrobacterium strains.

Benedict Reagent

Sodium Citrate 86.5 g
Sodium Carbonate, anhydrous 50 g

Add distilled water to make 425 mls

With rapid stirring, add 8.7 g Copper Sulfate (CuSO4  5H2O) in 50 mls distilled water

Library Freezing Mix (Texas A&M BAC Center)

The colonies are stored in LB with 10% Freezing Mix and the appropriate antibiotic.

10X Freezing Mix

360 mM K2HPO4
132 mM KH2PO4
17 mM NaCitrate
4 mM MgSO4
68 mM (NH4)2SO4
44% glycerol

We autoclave the Freezing Mix and store at room temperature. It is normal to see some precipitation after storage. We just mix well before alliquoting into the LB. We have not had any problems doing this. The freezing mix-LB-antibiotic mixture should be prepared fresh every few days.

Pipet 50 uls of LB + Freezing Mix + antibiotic per well with a multi-channel pipettor.

[BIBAC] Electroporation:
CM Hamilton lab protocols

Competent cells
DH10B, prepared as described by BioRad, or Gibco BRL “Electromax” cells.
The “electromax” cells are variable (for efficiency and contaminants) and often contain a low level contaminant that grows (slowly) on Kan+Sucrose.

BioRad Gene Pulser
200 OHMS
capacitance 25 microFD
Voltage depending on cuvettes – see below

Standard cloning: BioRad (or other) 0.2 cm gap; 2.5 kV; 40 microliters cells

Library construction (highest efficiency):
BioRad 0.1 cm gap; 1.8 kV
22 microliters cells

Mix DNA with aliquot of competent cells thawed on ice. Load cells + DNA into cuvette on ice.

Electroporate, do not put cuvette back on ice. Should get time constant of 4.5 – 4.8 (4.0-4.2 with library construction conditions). Remove cells from cuvette with microcapillary pipette tip and put into 400 microliters (standard transformation protocol) or 1 ml (for library construction) of SOC. Incubate at 37 degrees C for 45 minutes (libraries) to about 1 hour (standard cloning).

Selection for Inserts in BIBAC2 (37 degrees C)
Lennox Broth Agar + 40 mg/L kanamycin + 10% sucrose [with electromax cells we use 50 mg/L kanamycin}

Example: 500 mls agar + 50 mls of 50% sucrose autoclaved separately + 440 microliters of kanamycin (40 mg/ml kanamycin

Competent cells

Mersereau, M., Pazour, G.J. and Das, A. 1990. Efficient transformation of Agrobacterium tumefaciens by electroporation. Gene 90, 149-151. (We have an “in house” protocol based on this).

Electroporation protocol same as for E. coli – except recover for 1 hr at 30 degrees C. We typically use the 0.2 cm gap, 2.5V and 40 microliters of cells.

Selection for BIBAC2 and derivatives (30 degrees C)
LB + Kan 50 mg/L
Selection for pCH30/32/42
LB + Tet 5mg/L in C58 background, e.g. UIA143, GV3101
LB + Tet 2 mg/L in Ach5 background, e.g. LBA4404

Isolation of Plasmid DNA:

Ish -Horowicz and Burke (1981) Nuc. Acid Res. 9 (13):2989-2998

Grow 10 ml of desired E. coli strain in Luria broth plus appropriate antibiotic, at 37 degrees C shaking overnight.

Transfer 10 ml medium to 1 liter LB plus appropriate antibiotic in a 2 liter flask, shaking overnight 200-250 rpm.

Harvest cells in 500 ml bottles by centrifuging at 8000 rpm for 10 min. Using a Beckman JA10 or equivalent rotor this is 11,300g. Pour off supernatant. Add more overnight culture to the same bottle and spin again. The idea is to end up with one liter’s worth of pelleted cells in one 500 ml bottle.

Resuspend cells in 40 ml per liter culture of solution I. For single-copy plasmids: Add 10 ml solution I first, resuspend the pellet, add 25 ml solution I, then add 5 ml solution I with 200 mg Lysozyme, for a final concentration of 5 mg/ml lysozyme. After mixing in lysozyme, incubate on ice 10 min.

Add 80 ml per liter culture of solution II. Swirl and incubate on ice 10 minutes.

Add 40 ml per liter culture of cold solution III. Invert the bottle firmly but gently several times to mix. A sort of cottage-cheese like whit flock will form. Incubate on ice 15 minutes.

Add 10 ml of ddH2O per liter. Mix gently, and spin at 8000rpm; 20 minutes. Pour off supernatant though cheesecloth into 250 ml graduated cylinder. Record the volume of the supernatant, then pour it into a clean 500 ml centrifuge bottle.

Add 2-propanol equal to 0.6 volumes of the supernatant into the 500 ml bottle. Mix gently, by inversion and incubate on ice 5 minutes. Spin 8000rpm; 10 minutes.

Pour off supernatant and remove any remaining liquid with Pasteur pipette. Partially dry the pellet by passing air thought the bottle, or by storing the open bottle under a vacuum upside down.

The following works well for vTi65.2 rotor tubes which hold 5 mls total.  You want each pellet in a final volume 4 ml TE for VTi65.2 rotor. To do this add a small amount of TE to begin resuspending the pellet and adding 2M Tris-base dropwise with a pasteur pipette to begin to neutralization. Neutralize with 2M Tris-base, to pH 7.5-8.0, checking with pH paper. Measure the volume, and adjust to 4 mls.

For 4 ml volume add 4.2g Cesium Chloride (4 ml plus 0.2 ml EtBr which will be added later). After the CsCl has gone into the solution, add 0.2 ml of stock solution: 10 mg/ml Ethidium Bromide.

Centrifuge overnight at 55,000 rpm in vTi65.2, 25 degrees C

Visualize the DNA in by illuminating the gradient with a hand-held long UV wavelength (366 nm). If two bands are visible, the lower one is plasmid DNA. Sometimes with single-copy plasmid preps you will only see one band at this step. (With a high copy plasmid this might also be the case if the gradient is overloaded). Pull the plasmid band with an 18 gauge needle/ 3 ml syringe.

Place this DNA into a new ultracentrifuge tube, and then fill the tube up to the shoulder with CsCl/TE to make the gradient (e.g. 10g of CsCl added to 10 ml of TE). 50 microliters of 10mg/ml EthBr can also be added at the end without hurting the gradient.

Centrifuge again at 55,000 rpm for at least 6 hours. We routinely spin overnight.

Pull the lower plasmid DNA band and place into a small glass tube, or eppindorf tube.

Extract EtBr with TE-saturated butanol, or TE/CsCl-saturated 2-propanol

Dialyze Plasmid DNA against TE buffer at 4 degrees C with changes.

Solution I (store at 4 degrees C)
20% glucose 22.5 ml
1.0 M Tris pH = 8.0 12.5 ml
0.2 M EDTA pH = 8.0 25.0 ml
ddH2O q.a. 500 ml
Solution II (store at room temperature)
20% SDS 25.0 ml
5M NaOH 20.0 ml (add last)
ddH20 q.a. 500 ml
Solution III (store at 4 degrees C)
5M potassium acetate 300 ml
glacial acetic acid 200 ml
500 ml

Notes: Directions are for single or low copy plasmid, for ColE1or pUC type plasmids, 1 liter prep. should divided at least into 2 tubes.

For making vector for library construction, we typically prep 6 liters.

For plasmids larger than about 130 kb, recovery is reduced (probably due to damage/nicking that results in the loss of supercoils, which is the basis of the separation of plasmid from chromosomal DNA) so we usually prep 2L.

When preparing more than one liter – we band one liter per centrifuge tube, then combine (all) the bands into one tube for the second banding.

We have not been able to separate a 250 kb plasmid from chromosomal DNA by this method. Anne Frary purified a 250 kb BAC using Qiagen columns and the Qiagens modified protocol for isolating high molecular weight plasmids.

Agrobacterium plasmid prep:

Modified from P.J.J. Hooykaas and T. Mozo Plant Molecular Biology Manual B3: 1-9, 1994
CMH 5/95
Revised 9/95

Following this protocol, large single copy plasmids (including Ti plasmids) will be isolated as well as any higher copy and/or smaller plasmids. Since in most cases the binary plasmid is higher copy than the Ti, the restriction pattern of the binary will be brighter than the Ti pattern. In the case of a single copy binary plasmid, it may be necessary to do a Southern in order to resolve the pattern of the binary from the Ti.

In any case you will see bands that have come from digestion of the Ti plasmid. Therefore it is essential that you also prep the agrobacterium strain without the binary as a control. If you want to see a nice pattern for the pTi, in my experience NotI works nicely. I have not come up with a method that will CONSISTENTLY isolate small plasmids, but not large ones, so I just use this for everything.

This protocol works very well for C58 strains. Recovery is not as good for Ach5 strains (e.g., LBA4404).

  1. Grow 5 ml overnight LB + drug 36-48 hrs at 30 degrees C.
  2. Spin down cultures at about 1500g for 15 min at 4 degrees C (3500 rpm Beckman JA-10 rotor).
  3. Pour off supernatant and keep cultures on ice until all are ready to be resuspended.
  4. Resuspend pellets in 200 microliters solution I (50mM glucose, 10mMEDTA, 25mM TrisCl pH=8, 5 mg/ml lysozyme added fresh) by vortexing. When all cultures are resuspended, remove each to a 1.5 ml microfuge tube at room temp. Incubate 10 minutes.
  5. Add 400 microliters fresh solution II (1%SDS, 0.2N NaOH) invert tubes 4X and incubate 10 minutes at room temp. Invert tubes all at once by placing another microfuge rack on top to hold the tubes in place.
  6. Add 60 microliters of fresh alkaline phenol (2 volumes 0.2N NaOH + 1 volume Tris-saturated phenol mixed fresh before using – in my hands this is a single phase) mix by inverting 16X as above.
  7. Immediately add 300 microliters of 3M NaOAc pH 5.0, mix by inverting 20X.
  8. Incubate at —20 degrees C for 20 minutes
  9. Centrifuge 5 minutes at room temp using a microfuge that has slow acceleration and slow deceleration options (such as the “turtle” setting on an IEC MicroMax).
  10. Collect the supernatant and put into a 2.2 ml microfuge tube. If the inversions have been done properly you will be able to remove about 800 microliters. Sometimes there is still a lot of flock and you will only remove about 600. I remove the same amount from each tube.
  11. Add an equal volume of Tris-equilibrated phenol, and extract by inverting 20X.
  12. Spin 5 min at room temp “turtle”.
  13. Pipette off upper phase into a 2.2 ml microgfuge tube and add two volumes of ice cold (-20 degrees) 95% ethanol.
  14. Mix by inversion 4X and spin 10 min at room temp “turtle”.
  15. Wash pellet with 500 microliters ice cold (-20 degrees) 70% ethanol, spin 2 minutes “turtle”.
  16. Dry (approximately) 10 minutes in a vacuum desicator.
  17. Resuspend (gently) in 40 microliters TE.
  18. Cut 15 microliters for a standard gel, 20 microliters for a pulse-field gel.

Gel conditions: 0.7% agarose, 1X TAE buffer 30-40V overnight, or pulsed-field gel.

NOTE:  If you “pulse down” the finished preps you must use slow acceleration/deceleration. You should also keep pipetting to a minimum. However, I have not found it necessary to use wide bore tips for pipetting to set up digests and load gels etc.

Preparation of Agrobacterium electroporation competent cells:

Ref: Mersereau et al. (1990) Gene 90, 149-151.
C. M. Hamilton lab protocol

Inoculate 500 ml of Lennox broth (Gibco BRL) with 5 ml of an “overnight” (24-36 hrs) saturated culture in a 2 Liter flask. Shake overnight 250 rpm, at 30 degrees C. [We usually start the cultures around 3 or 4 PM, and harvest around 9 AM]. Harvest culture at A 600 = 1.5 (1.2 is OK).

Spin down cells at 5000 rpm 10 minutes in sterile 500 ml bottles. Drain well. Wash 4-5X with equal volume of Milli-Q grade water, cold and sterile.

For each wash, add a small volume 5-10 mls of water to the cells and resuspend cells by pipetting gently until no clumps remain. Do this gently with a wide bore sterile pipette. Then add the rest of the water and mix gently.

After the 4th or 5th wash, resuspend cells in 10 mls water and transfer to sterile (we use screw cap) 15 ml centrifuge tubes. Spin at 6000 rpm for 10 minutes. Resuspend cells pellet by adding 1-2 mls 10% glycerol. You want the total volume of resuspended cells to be 2.5-4 mls depending on the OD that you harvested at.

Aliquot 40 microliters into sterile microfuge tubes on dry ice.

Store at -70 degrees C.

Final concentration of cells is 4-6 X 10 to the tenth.

Materials Needed
2L flask with 500 ml LB [We use hand made cheesecloth/cotton tops covered with a piece of foil to seal the flask]. For (1) 500 ml culture: about 3 Liters of sterile cold Milli-Q or other good quality (distilled and deionized) water. Also sterile centrifuge bottles/tops, and 10% glycerol, cold and sterile. Dry ice, Sterile microfuge tubes.

Electroporation Conditions
We do our electroporations exactly as we do for E. coli, except that the cells recover at 30 degrees C for about one hour instead of 37 degrees C (E. coli). We use a BioRad Gene-Pulser and manufacturer recommended settings and protocol.

We plate out the cells on Lennox broth agar + 50 mg/L kanamycin to select for the BIBAC and 5 mg/L tetracycline for the pCH30 or pCH32 plasmid.

EXCEPT for LBA4404 strains the tet level is 2 mg/L for the helper plasmid. You should have colonies in 36 – 40 hrs.

Construction of tomato genomic DNA BIBAC libraries – detailed protocol:

Experimental procedures

Vector preparation
Approximately 10 &g of CsCl purified BIBAC2 plasmid DNA were digested with 30 unites of BamHI (GIBCO BRL, Grand Island, NY) in the reaction buffer supplied by the manufacturer with the addition of 2 mM (final concentration) spermidine (S2501, Sigma, St. Louis MO) in a volume of 200 &l at 37° C for 2 h, then adding 10 units for 1 h. Digested DNA was quantified on an agarose gel and subsequently treated with calf intestinal alkaline phosphatase (CIAP; GIBCO BRL) 0.1 unit per pmol of vector DNA, for 30 min at 37° C. The reaction was stopped by adding 2 &l (1/100 volumes) 0.5 M EDTA, 5 &l (1/40 volume) 20% SDS and 20 &l (1/10 volume) 1 mg/ml Proteinase K (GIBCO BRL in distilled deionized water) and the reaction mix was incubated at 56% C for 30 min and allowed to cool at room temperature for 10 min. Then, the reaction mix was extracted with an equal volume of (1:1) phenol (equilibrated with Tris-Cl pH 8):chloroform. The aqueous phase was removed and the DNA precipitated by addition of 1/10 volume 3 M sodium acetate pH 7.0 and 2 volumes of 95% ethanol, followed by centrifugation and a 70% ethanol wash. The BIBAC2 vector DNA was resuspended in distilled deionized water at a final concentration of 40 – 50 ng/&l and frozen in aliquots at -20% C.

Preparation of tomato nuclei
Fifteen g of young tomato leaf tissue were harvested from greenhouse grown plants, quick frozen in liquid nitrogen, and stored at -80° C. The tissue was ground to a smooth powder using a mortar and pestle with several additions of liquid nitrogen during the grinding process. The plant tissue powder was added to 150 ml of (HB*) homogenization buffer (Zhang, et al. 1995) with 0.5% Triton-X100 and 0.15% beta-mercaptoethanol in a sterile beaker (on ice) stirring at low speed for about 20 minutes. The resuspended tissue was then filtered through two double layers of cheesecloth and a single layer of miracloth into a 250 ml centrifuge bottle. The nuclei were pelleted by centrifugation at 3500 rpm (2000 g) for 20 minutes at 4° C. The supernatant was decanted and the pellet was resuspended very gently (with a paintbrush) in HB* and half of the suspension was transferred to each of 2, 40 ml centrifuge tubes. Additional HB* was added to fill the 40 ml sterile centrifuge tubes (on ice). The nuclei were pelleted by centrifugation as before, and this wash step was repeated. The pellets were then combined into one 40 ml tube and processed as before. Finally the pellet was resuspended by adding approximately 11 ml of HB, for a final volume of 12 ml. Preparation of low-melting-point (LMP) agarose (Seaplaque GTG, FMC, Rockland, ME) beads or plugs was essentially as described by Zhang, et al., (1995). Lysis of the nuclei in the embedded agarose, and subsequent washes were carried out as described by Woo, et al., (1994).

Digestion and size selection of tomato genomic DNA
Prior to digestion with restriction endonuclease, the beads or plugs were washed (on ice) in the buffer recommended by the manufacturer (BRL React 3 for digestion with BamHI) with addition of 0.2 mg/ml BSA, 2 mM spermidine and 1 mM DTT (final concentrations). Three, one hour washes were carried out. Optimal partial conditions were determined for 100 &l of beads or one plug (plug mold from Bio-Rad, Hercules, CA) in a 200 &l reaction volume using 2-4 units of BamHI, checking the extent of digestion at several time points between 2 and 20 min by pulsed-field gel electrophoresis. Multiple partial digests were then made using the optimized conditions and were loaded into one large preparative well for pulsed-field gel electrophoresis. The pulsed-field gel conditions used for visualization of partial digests were: 6V/cm, 15 sec pulse (constant), 120 degree angle, 11°-14° C, 1% agarose, 0.5X TBE, for 15 h. For the preparative gel, the lower portion of the gel was made with reagent grade LMP agarose (FMC or Bio-Rad).

Several size classes of digested genomic DNA were chosen (approximately 50 -100 kb; 100-200 kb, 200-300 kb). The gel slices containing the size selected DNA were melted at 68° C for 10 min then each was loaded into a small preparative well on a 1% LMP gel and subjected to pulsed-field gel electrophoresis at 4V/cm, 5 sec pulse (constant), 120 degree angle, 11°-14° C, 0.5X TBE, for 6-8 h. The most concentrated portion of the DNA from each size class was cut from the gel, the agarose digested away with gelase (Epicentre Technologies, Madison, WI) quantified and used for ligation. Portions of the gels were stained with ethidium bromide for visualization. However, the DNA that was used for ligation was never subjected to staining with ethidium bromide. Pulsed-field gel electrophoresis was carried out using a CHEF-Mapper or DRIII apparatus (both Bio-Rad).

Ligation reaction mixes
The ligation reaction mixes used to construct these libraries contained approximately 600 ng of genomic DNA (100-200 kb size selection) and 200 ng of BIBAC2 vector (22.5 kb). The vector and genomic DNA were mixed together and heated at 60° C for 10 minutes then cooled to room temperature. Next, ligation buffer (10X) and T4 DNA ligase (1 unit/&l) (both USB, Cleveland, OH) were added to the DNA, mixing gently. The ligation mixes (approximately 200 &l) were then incubated overnight at 16° C. The following day the ligation mix was frozen in aliquots at -20° C.

If higher transformation efficiency was desired (more colonies recovered per electroporation), then an aliquot (20 &l) of ligation mix was dialyzed on a filter (Millipore, Bedford, MA) against sterile deionized distilled water at room temperature for 30 minutes. It was not possible to recover all of the original volume and there is some evaporative loss; however, we typically observed a 3-5 fold increase in electroporation efficiency after this dialysis step. It was also possible to add 2-3 &l of a dialyzed ligation mix to a single electroporation cuvettes without experiencing arcing. Increasing transformation efficiency per electroporation can help reduce the overall cost of generating the library.

Transformation, selection and analysis of clones
E. coli DH10B cells and a Gene Pulser (Bio-Rad) were used for electroporation. Electroporation competent E. coli cells for preliminary experiments were prepared according to the protocol provided by Bio-Rad. To generate the libraries, electroporation competent E. coli cells were purchased from GIBCO BRL (DH10B Electromax). Electroporation conditions were as recommended by the manufacturer for 0.1 cm gap cuvettes (Bio-Rad). The transformed cells were allowed to recover for 40 min in SOC media (Sambrook, et al. 1982) at 37°C, then spread onto Lennox agar (GIBCO BRL) plates containing kanamycin (80 mg/L) and sucrose (5%) (kanamycin and sucrose were added as sterile solutions just before pouring the plates). The plates were incubated for approximately 24 h at 37° C. Individual colonies were used to inoculate 5 ml liquid cultures and were grown in a rolling-wheel shaker for approximately 24 h at 37° C. Plasmid DNA was isolated (BAC mini-preps), digested with NotI (NEB, Beverly, MA) and visualized following pulsed-field electrophoresis as described by Woo et al., (1994).

To generate the libraries, individual colonies were picked into 384 well plates. Each well contained 50 &l of LB containing 10% freezing mix [36 mM K2HPO4, 13.2 mM KH2PO4, 1.7 mM sodium citrate, 0.4 mM MgSO4, 68 mM (NH4)2SO4, 4.4% glycerol (final concentrations)] with kanamycin (50 mg/L final concentration) added. The cultures were incubated at 37° C for approximately 24 h in an airtight container to prevent evaporative loss. The next day the plates were checked and any wells that did not grow were marked. The plates were then stored at -70° C.