Information

Why pellet and resuspend E. coli for plasmid prep


For maxipreps, cant you just add all the stuff that would be in P1 (RNase A, EDTA) then just add P2? Because if we pellet cells then resuspend theres got to be a reason, right? Is it slats and other stuff in the media that mess with lysis?


The reason this centrifugation/resuspension step is done has a simple reason: It concentrates the bacteria in a much smaller volume which is much easier to handle afterwards.

For a maxiprep you typically have a culture volume between 250 and 500ml, depending on the copy number of the plasmid. After centrifugation the pellet is resuspended in 10ml of Buffer P1, for lysis another 10ml of buffer P2 are added, for neutralization 10ml of P3.

Imagine doing this with 500ml culture volume - adding 500ml of each P2 and P3. This might work for a miniprep, but definitely not for larger volumes. Additionally, the centrifugation steps are carried out at high g numbers - typically something around 15.000xg. Doing this with this really high volumes is complicated and needs big centrifuges, which a lot of people are not trained anymore to operate or have available.


Purification and Identification of Plasmid DNA

1 Purification and Identification of Plasmid DNA LABORATORY 8 GROWTH OF E. COLI ON AMPICILLIN PLATES demonstrates transformation to an antibiotic-resistant phenotype. In the basic version of Laboratory 5, the observed phenotype was due to uptake of plasmid pamp, a DNA molecule that is wellcharacterized. In experimental situations where numerous recombinant plasmids are generated by joining two or more DNA fragments, the antibiotic resistance marker only functions to indicate which cells have taken up a plasmid bearing the resistance gene. It does not indicate anything about the structure of the new plasmid. Therefore, it is standard procedure to isolate plasmid DNA from transformed cells and to identify the molecular genotype using DNA restriction analysis. In cases where the recombinant molecules are formed by combining well-characterized fragments, restriction analysis is sufficient to confirm the structure of a hybrid plasmid. In other cases, the nucleotide sequence of the insert must be determined. This protocol is divided into two parts: Plasmid Minipreparation of pamp and Restriction Analysis of Purified pamp. Part A provides a small-scale protocol to purify from transformed E. coli enough plasmid DNA for restriction analysis. Cells taken from an ampicillinresistant colony are grown to stationary phase in suspension culture. The cells from 1 ml of culture are harvested and lysed, and plasmid DNA is separated from the cellular proteins, lipids, and chromosomal DNA. This procedure yields 2 5 µg of relatively crude plasmid DNA, in contrast to large-scale preparations that yield 1 mg or more of pure plasmid DNA from a 1-liter culture. Part B provides a protocol using a sample of plasmid DNA isolated in Part A and a control sample of pamp. These two samples are cut with the restriction enzymes BamHI and HindIII and coelectrophoresed on an agarose gel, and the restriction patterns are stained and visualized. The purified DNA is shown to have a restriction fingerprint identical to that of pamp. Bam/Hind restriction fragments of the miniprep DNA comigrate with the 784-bp and 3755-bp Bam/Hind fragments of pamp. This provides genotypic proof that pamp molecules were successively transformed into E. coli in Laboratory 5. Kits based on this laboratory are available from the Carolina Biological Supply Company. Catalog no (with ethidium bromide stain) Catalog no (with the CarolinaBlu stain) 423

2 424 LABORATORY 8 PART A Plasmid Minipreparation of pamp ADD to 2 tubes CENTRIFUGE POUR OFF supernatant DRAIN (2 min) E. coliipamp ADD to both tubes RESUSPEND ADD and MIX INCUBATE (5 min) GTE SDS/NaOH ADD and MIX both tubes INCUBATE (5 min) CENTRIFUGE (5 min) TRANSFER supernatant KOAc ADD and MIX both tubes CENTRIFUGE POUR OFF supernatant DRAIN (5 min) Isopropanol ADD and FLICK both tubes CENTRIFUGE POUR OFF supernatant DRAIN (3 min) Ethanol DRY both tubes ADD RESUSPEND POOL (5 min) TE DNA/TE

3 PURIFICATION AND IDENTIFICATION OF PLASMID DNA 425 PRELAB NOTES Optimally, minipreps should be done on cells that have been recently manipulated for transformation. This completes a conceptual stream that firmly cements the genotype-phenotype relationship. Alternatively, use streaked plates of transformed E. coli to prepare overnight cultures. Plasmid Selection pamp gives superior yields on minipreps compared to pbr322. A derivative of a puc expression vector, pamp is highly amplified more than 100 copies are present per E. coli cell. If substituting a different plasmid for miniprep purposes, select a commercially available member of the puc family, such as puc18 or puc19. Centrifuge Requirements A microfuge that generates approximately 12,000 times the force of gravity (12,000g) is needed for efficient and rapid purification of plasmid DNA. A slower-spinning clinical or preparatory centrifuge cannot be substituted. Supplies Sterile supplies are not required for this protocol. Standard 1-ml pipettes, transfer pipettes, and/or microcapillary pipettes can be used instead of micropipettors. Use good-quality, colorless 1.5-ml tubes, beginning with Step 11. The walls of poor-quality tubes, especially colored tubes, often contain tiny air bubbles that can be mistaken for ethanol droplets in Step 19. We have observed students drying DNA pellets for 15 minutes or more, trying to rid their tubes of these phantom droplets. (Typical drying time is actually several minutes.) Fine Points of Technique Be careful not to overmix reagents excessive manipulation shears chromosomal DNA. The success of this protocol in large part depends on maintaining chromosomal DNA in large pieces that can be differentially separated from intact plasmid DNA. Mechanical shearing increases the amount of short-sequence chromosomal DNA, which is not removed in the purification of plasmid DNA. Make sure that the microfuge will be immediately available for Step 13. If sharing a microfuge, coordinate with other experimenters to begin Steps 12 and 13 together. For Further Information The protocol presented here is based on the following published methods: Birnboim H.C. and Doly J A rapid alkaline extraction method for screening recombinant plasmid DNA. Nucleic Acids Res. 7: Ish-Horowicz D. and Burke J.F Rapid and efficient cosmid cloning. Nucleic Acids Res. 9:

4 426 LABORATORY 8 PRELAB PREPARATION Before performing this Prelab Preparation, please refer to the cautions indicated on the Laboratory Materials list. 1. This protocol is designed to follow Laboratory 5. Ideally, students should pick colonies from their own transformed plates to begin this experiment. However, the colonies must grow overnight or for at least several hours, so unless the class meets on consecutive days, it may be necessary for the instructor to set up the cultures used in this laboratory. In any case, on the day before the laboratory, prepare an E. coli culture according to the protocol in Laboratory 2B, Overnight Suspension Culture. Inoculate the culture with a cell mass scraped from one colony selected from the +LB/amp plate from Laboratory 5, Rapid Colony Transformation of E. coli with Plasmid DNA. Maintain antibiotic selection with LB broth plus ampicillin. Alternatively, prepare the culture 2 3 days in advance and store at 4ºC or incubate for hours at 37ºC without shaking. In either case, the cells will settle at the bottom of the culture tube. Shake the tube to resuspend cells before beginning procedure. 2. The SDS/sodium hydroxide solution should be fresh prepare this solution within a few days of lab. Store solution at room temperature a soapy precipitate may form at lower temperature. If a precipitate forms, warm the solution by placing the tube in a beaker of hot tap water, and shake gently to dissolve the precipitate. 3. Prepare aliquots for each experiment: 250 µl of glucose/tris/edta (GTE) solution (store on ice) 500 µl of SDS/sodium hydroxide (SDS/NaOH) solution 400 µl of potassium acetate/acetic acid (KOAc) solution (store on ice) 1000 µl of isopropanol 500 µl of 95% ethanol 50 µl of Tris/EDTA (TE) solution 4. Review Part B, Restriction Analysis of Purified pamp.

5 PURIFICATION AND IDENTIFICATION OF PLASMID DNA 427 MATERIALS CULTURES AND MEDIA E. coli/pamp overnight culture Ethanol (95 100%) Glucose/Tris /EDTA (GTE) Isopropanol Potassium acetate/acetic acid (KOAc) SDS/sodium hydroxide (SDS/NaOH) Tris /EDTA (TE) SUPPLIES AND EQUIPMENT Beakers for crushed ice and for waste/used tips Bleach (10%) or disinfectant Clean paper towels Hair dryer Microfuge Micropipettor ( µl and µl) + tips Permanent marker Test tube rack Tubes (1.5-ml) See Appendix 4 for Cautions list. METHODS The cell pellet will appear as a small off-white smear on the bottom-side of the tube. Although the cell pellets are readily seen, the DNA pellets in Step 14 are very difficult to observe. Make a habit of aligning the tube with the cap hinges facing outward in the microfuge rotor. Then, pellets should always be located at the tube bottom beneath the hinge. Accurate pipetting is essential to good plasmid yield. The volumes of reagents are precisely calibrated so that the sodium hydroxide added in Step 6 is neutralized by acetic acid in Step 8. Prepare Duplicate Minipreps (50 minutes) The instructions below are for making duplicate minipreps, which provide balance in the microfuge and insurance if a critical mistake is made. 1. Shake culture tube to resuspend E. coli cells. 2. Label two 1.5-ml tubes with your initials. Use a micropipettor to transfer 1000 µl of E. coli/pamp overnight suspension into each tube. 3. Close caps, and place the tubes in a balanced configuration in the microfuge rotor. Spin for 1 minute to pellet cells. 4. Pour off supernatant from both tubes into a waste beaker for later disinfection. Alternatively, use a micropipettor to remove supernatant. Be careful not to disturb the cell pellets. Invert the tubes, and tap gently on the surface of a clean paper towel to drain thoroughly. 5. Add 100 µl of ice-cold GTE solution to each tube. Resuspend the pellets by pipetting the solution in and out several times. Hold the tubes up to the light to check that the suspension is homogeneous and that no visible clumps of cells remain. 6. Add 200 µl of SDS/NaOH solution to each tube. Close caps, and mix solutions by rapidly inverting tubes five times. 7. Stand tubes on ice for 5 minutes. Suspension will become relatively clear. 8. Add 150 µl of ice-cold KOAc solution to each tube. Close caps, and mix solutions by rapidly inverting tubes five times. A white precipitate will immediately appear.

6 Step 11, the supernatant is ved and precipitate is disrded. The situation is reverin Steps 14 and 17, where e precipitate is saved and e supernatant is discarded. Step 12 quickly, and make re that the microfuge will immediately available for p 13. e pellet may appear as a y smear or small particles the bottom-side of each e. Do not be concerned if llet is not visible pellet size not a predictor of plasmid ld. A large pellet is comsed primarily of RNA, and lular debris carried over m the original precipitate. smaller pellet often means leaner preparation. cleic acid pellets are not solle in ethanol and will not uspend during washing. to expelling supernatant. Discard old tubes containing precipitate. 12. Add 400 &mul of isopropanol to each tube of supernatant. Close caps, and mix vigorously by rapidly inverting tubes five times. Stand at room temperature for only 2 minutes. (Isopropanol preferentially precipitates nucleic acids rapidly however, proteins remaining in solution also begin to precipitate with time.) 13. Place tubes in a balanced configuration in the microfuge rotor, and spin for 5 minutes to pellet the nucleic acids. Align tubes in rotor so that cap hinges point outward. The nucleic acid residue, visible or not, will collect on the tube side under the hinge during centrifugation. 14. Pour off supernatant from both tubes. Be careful not to disturb nucleic acid pellets. Alternatively, remove the supernatant with a 1000-ml micropipettor. Place tip away from the pellet. If you are concerned that the pellet has been drawn up in the tip, transfer the supernatant to another 1.5-ml tube, recentrifuge, and remove the supernatant again. Invert tubes, and tap gently on the surface of a clean paper towel to drain thoroughly. 15. Add 200 &mul of 100% ethanol to each tube, and close caps. Flick tubes several times to wash pellets. Store DNA in ethanol at 20ºC until ready to continue. 16. Place tubes in a balanced configuration in microfuge rotor, and spin for 2 3 minutes. 17. Pour off supernatant from both tubes. Be careful not to disturb nucleic acid pellets. Alternatively, remove the supernatant with a 1000-ml micropipettor. Place tip away from the pellet. If you are concerned that the pellet has been drawn up in the tip, transfer the supernatant to another 1.5-ml tube, recentrifuge, and remove the supernatant again. Invert tubes, and tap gently on the surface of a clean paper towel to drain thoroughly. 18. Dry nucleic acid pellets by one of the following methods:

7 21. Pool DNA/TE solution into one tube. Freeze DNA/TE solution at 20ºC until ready to continue. Thaw before using. 22. Take time for responsible cleanup. a. Segregate for proper disposal culture tubes and micropipettor tips tha have come in contact with E. coli. b. Disinfect overnight culture, tips, and supernatant from Step 4 with 10% bleach or disinfectant. c. Wipe down lab bench with soapy water, 10% bleach solution, or disin fectant (such as Lysol). d. Wash hands before leaving lab. RESULTS AND DISCUSSION The minipreparation is a simple and efficient procedure for isolating plasmi DNA. Become familiar with the molecular and biochemical effects of eac reagent used in the protocol. Glucose//Tris/EDTA: The Tris buffers the cells at ph 7.9. EDTA binds divalen cations in the lipid bilayer, thus weakening the cell envelope. SDS/sodium hydroxide: This alkaline mixture lyses the bacterial cells. The deter gent SDS dissolves the lipid components of the cell membrane, as well as ce lular proteins. The sodium hydroxide denatures the chromosomal and plas mid DNA into single strands. The intact circles of plasmid DNA remain inter twined. Potassium acetate/acetic acid: The acetic acid returns the ph to neutral, allowin DNA strands to renature. The large, disrupted chromosomal strands canno rehybridize perfectly, but instead collapse into a partially hybridized tangle. A

8 430 LABORATORY 8 a higher vapor point than does ethanol. The ethanol-isopropanol evaporates more rapidly in the drying step. Tris/EDTA: Tris buffers the DNA solution. EDTA protects the DNA from degradation by DNases by binding divalent cations that are necessary cofactors for DNase activity. Buffering DNA is important, as low ph (<6) leads to the loss of purines (adenine and guanine) called depurination. The purines are actually cleaved from their sugars, creating an abasic site. Purine cleavage is a very common occurrence in cells (on the order of 10 5 times per cell per day) and is repaired by specific DNA repair systems. Of course, your DNA is in a tube and there is no DNA repair system present to repair it. Keep in mind that H 2 O can have a ph as low as Consider the three major classes of biologically important molecules: proteins, lipids, and nucleic acids. Which steps of the miniprep procedure act on proteins? On lipids? On nucleic acids? 2. What aspect of the plasmid DNA structure allows it to renature efficiently in Step 8? 3. What other kinds of molecules, in addition to plasmid DNA, would you expect to be present in the final miniprep sample? How could you find out? FOR FURTHER RESEARCH Determine the approximate mass of plasmid DNA you isolated per milliliter of cells. 1. Set up 20-µl HindIII restriction reactions using 15 µl of your pamp preparation and a known mass of &lambda DNA as a control. 2. Make 1:10, 1:50, and 1:100 dilutions of the digested pamp and &lambda DNAs. 3. Separate by electrophoresis equal volumes of each dilution in an agarose gel, and stain with ethidium bromide. CAUTION Review Responsible Handling of Ethidium Bromide in Laboratory 3. Wear latex gloves when staining, viewing, and photographing gel and during cleanup. Confine all staining to a restricted sink area. For further information, see Appendix Identify a lane of the &lambda digest where the 4361-bp fragment is just visible, and identify a lane of pamp (4539 bp) that is of equal intensity. These bands should have a nearly equivalent mass of DNA. 5. Determine the mass of &lambda DNA in the selected fragment, using the formula below. Make sure to account for the dilution factor. fragment bp (conc. DNA) (vol. DNA) &lambda bp 6. Multiply the mass from Step 5 by the dilution factor of the selected pamp lane.

9 PURIFICATION AND IDENTIFICATION OF PLASMID DNA 431 PART B Restriction Analysis of Purified pamp I. Set Up Restriction Digest ADD Mini Mini+ pamp+ pamp Miniprep DNA Buf/RNase H 2 O Miniprep DNA Buf/RNase Bam/Hind H 2 O pamp Buf/RNase Bam/Hind H 2 O pamp Buf/RNase H 2 O MIX all tubes INCUBATE all tubes 37 C II. Cast 0.8% Agarose Gel POUR gel SET III. Load Gel and Separate by Electrophoresis ADD to all tubes Loading dye LOAD gel ELECTROPHORESE volts + IV. Stain Gel and View (Photograph) STAIN gel RINSE gel VIEW gel PHOTOGRAPH gel

10 432 LABORATORY 8 PRELAB NOTES Review Prelab Notes in Laboratory 3, DNA Restriction Analysis. Limiting DNase Activity Unlike highly purified plasmid DNA available from commercial vendors, miniprep DNA is impure. A significant percentage of nucleic acid in the preparation is, in fact, RNA and fragmented chromosomal DNA. Typically, miniprep DNA is contaminated with nucleases (DNases) that cleave DNA into small fragments. Residual DNases will degrade plasmid DNA if minipreps are left for long periods of time at room temperature or even on ice. For this reason, store minipreps at 20ºC, and thaw just prior to use. The situation is further complicated during restriction digestion. DNases and restriction endonucleases both require divalent cations, such as Mg ++. Included in TE buffer at a low concentration of 1 mm, Na 2 EDTA chelates (binds) divalent cations at a ratio of 2 cations/na 2 EDTA. Thus, a 1 mm solution can chelate about 2 mm of divalent cation. Although some divalent cations may remain free, we are limited to how much Na 2 EDTA can be added because higher concentrations would chelate the Mg ++ necessary for restriction enzyme activity. A balance is thus struck at an EDTA concentration that inhibits most of the contaminating DNases without significantly reducing the activity of the restriction enzymes. Another balance must be struck. On the one hand, contaminants in the miniprep limit restriction enzyme activity a 20-minute incubation is not usually sufficient for complete digestion. On the other hand, DNases are activated by Mg ++ in the restriction buffer and will significantly degrade plasmid DNA if the restriction reaction is incubated too long. Experience has shown that a 30- minute incubation gives optimal results. RNase Miniprep DNA is contaminated by large amounts of ribosomal RNA and smaller amounts of messenger RNA and transfer RNA. If not removed from the preparation, this RNA will obscure the DNA bands in the agarose gel. Therefore, RNase is added to the restriction digest during incubation, the RNase digests RNA into very small fragments (less than 100 nucleotides). These RNA fragments run well ahead of the DNA fragments of interest or are so small that they do not stain. For Further Information The protocol presented here is based on the following published methods: Aaij C. and Borst P The gel electrophoresis of DNA. Biochim. Biophys. Acta 269: Helling R.B., Goodman H.M., and Boyer H.W Analysis of R-EcoRI fragments of DNA from lambdoid bacteriophages and other viruses by agarose-gel electrophoresis. J. Virol. 14: Sharp P.A., Sugden B., and Sambrook J Detection of two restriction endonuclease activities in Haemophilus parainfluenzae using analytical agarose-ethidium bromide electrophoresis. Biochemistry 12:

11 PURIFICATION AND IDENTIFICATION OF PLASMID DNA 433 PRELAB PREPARATION MATERIALS Before performing this Prelab Preparation, please refer to the cautions indicated on the Laboratory Materials list. 1. Mix in 1:1 proportion: BamHI + HindIII (6 µl per experiment). Keep on ice. 2. Prepare aliquots for each experiment: 12 µl of 0.1 µg/µl pamp (store on ice) 12 µl of 5x restriction buffer/rnase (store on ice) 6 µl of BamHI/HindIII (store on ice) 500 µl of distilled water 500 µl of loading dye If another plasmid was substituted for pamp in the transformation, use that plasmid as a control in the restriction digest. 3. Prepare 0.8% agarose solution (

40 50 ml per experiment). Keep agarose liquid in a hot-water bath (at

60ºC) throughout the experiment. Cover the solution with aluminum foil to retard evaporation. 4. Prepare 1x Tris/Borate/EDTA (TBE) buffer for electrophoresis ( ml per experiment). 5. Prepare ethidium bromide or methylene blue staining solution (100 ml per experiment), or other proprietary stain. 6. Adjust water bath to 37ºC. REAGENTS Agarose (0.8%) BamHI/HindIII (50:50 mix) Distilled water Ethidium bromide (1 µg/ml) (or 0.025% methylene blue ) Loading dye Miniprep DNA pamp (0.1 µg/µl) 5x Restriction buffer/rnase 1x Tris /Borate/EDTA (TBE) buffer SUPPLIES AND EQUIPMENT Aluminum foil Beakers for agarose and for waste/ used tips Camera and film (optional) Electrophoresis box Latex gloves Masking tape Microfuge (optional) Micropipettor ( µl) + tips Parafilm or wax paper (optional) Permanent marker Plastic wrap (optional) Power supply Test tube rack Transilluminator (optional) Tubes (1.5-ml) Water baths (37ºC and 60ºC) See Appendix 4 for Cautions list.

12 2. Use the matrix below as a checklist while adding reagents to each reaction. Read down each column, adding the same reagent to all appropriate tubes. Mini = miniprep, no enzymes Mini+ = miniprep + BamHI/HindIII pamp+ = pamp + BamHI/HindIII pamp = pamp, no enzymes Use a fresh tip for each reagent. Refer to detailed directions that follow. Miniprep Buffer/ BamHI/ Tube DNA pamp RNase HindIII H 2 O 37 C o not overincubate. During nger incubation, DNases in e miniprep may degrade asmid DNA. 3. Collect reagents, and place in test tube rack on lab bench (BamHI/HindIII on ice). 6. Use a fresh tip to add 2 &mul of restriction buffer/rnase to a clean spot on each 7. Use a fresh tip to add 2 &mul of BamHI/HindIII to tubes labeled Mini+ and 9. Close tube tops. Pool and mix reagents by pulsing in a microfuge or by 10. Place reaction tubes in a 37ºC water bath, and incubate for 30 minutes only. Mini 5 &mul 2 &mul 3 &mul Mini+ 5 &mul 2 &mul 2 &mul 1 &mul pamp+ 5 &mul 2 &mul 2 &mul 1 &mul pamp 5 &mul 2 &mul 3 &mul 4. Add 5 &mul of miniprep DNA to tubes labeled Mini and Mini+. 5. Use a fresh tip to add 5 &mul of pamp to tubes labeled pamp+ and pamp. reaction tube. pamp+. 8. Use a fresh tip to add the proper volumes of distilled water to each tube. sharply tapping the tube bottom on the lab bench. Following incubation, freeze reactions at 20ºC until ready to continue. Thaw reactions before continuing to Section III, Step 1.

13 buffer remaining in electrophoresis box from a previous experiment, rock chamber back and forth to remix ions that have accumulated at either end. Buffer solution helps to lubricate comb. Some gel boxes are designed such that the comb must be removed prior to inserting casting tray into box. In this case, flood casting tray and gel surface with running buffer before removing comb. Combs removed from a dry gel can cause tearing of wells. A piece of dark construction paper beneath the gel box will make the wells more visible. 6. Gently remove comb, taking care not to rip wells. 7. Make certain that sample wells left by the comb are completely submerged If dimples appear around the wells, slowly add buffer until they disappea Cover the electrophoresis tank and save the gel until ready to continue. Ge will remain in good condition for at least several days if it is completely submerged in buffer. III. Load Gel and Separate by Electrophoresis (30 50 minute 1. Add loading dye to each reaction. Either a. Add 1 &mul of loading dye to each reaction tube. Close tube tops, and mi by tapping the tube bottom on the lab bench, pipetting in and out, o pulsing in a microfuge. Make sure that the tubes are placed in a balance configuration in the rotor. or b. Place four individual droplets of loading dye (1 &mul each) on a small squar of Parafilm or wax paper. Withdraw contents from reaction tube, and mi with a loading dye droplet by pipetting in and out. Immediately load dy mixture according to Step 2. Repeat successively, with a clean tip, for eac reaction. 2. Use a micropipettor to load 10 &mul of each reaction tube into a separate we in the gel, as shown on the following page. Use a fresh tip for each reaction a. Before loading sample, make sure that there are no bubbles in the well If bubbles exist, remove them with a micropipettor tip. b. Use two hands to steady the micropipettor over the well. c. If there is air in the end of the tip, carefully depress the plunger to push th sample to the end of the tip. (If an air bubble forms a cap over the we DNA/loading dye will flow into the buffer around the edges of the well.)

14 ternatively, set power pply on lower voltage, d run gel for several urs. When running two ls from the same power pply, current is double at for a single gel at the me voltage. e BamHI/HindIII digest lds two bands containing all fragments of 784 bp d 3755 bp, which are sily resolved during a ort electrophoresis run. e 784-bp fragment runs ectly behind the purplish nd of bromophenol blue uivalent to

300 bp), ereas the 3755-bp fragnt runs in front of the ua band of xylene cyanol uivalent to

9000 bp). 4. Turn power supply on, and set to volts. The ammeter should register approximately milliamperes. If current is not detected, check connections and try again. 5. Separate by electrophoresis for minutes. Good separation will have occurred when the bromophenol blue band has moved 4 7 cm from the wells. If time allows, carry out electrophoresis until the bromophenol blue band nears the end of the gel. Stop electrophoresis before the bromophenol blue band runs off the end of the gel. 6. Turn off power supply, disconnect leads from the inputs, and remove the top of the electrophoresis box. 7. Carefully remove the casting tray from the electrophoresis box, and slide the gel into a disposable weigh boat or other shallow tray. Label staining tray with your name. Cover the electrophoresis tank and save the gel until ready to continue. Gel can be stored in a zip-lock plastic bag and refrigerated overnight for viewing/photographing the next day. However, over longer periods of time, the DNA will diffuse through the gel and the bands will become indistinct or disappear entirely. ining may be performed an instructor in a conlled area when students e not present. 8. Stain and view gel using one of the methods described in Sections IVA and IVB. IVA. Stain Gel with Ethidium Bromide and View (Photograph) (10 15 minutes) CAUTION Review Responsible Handling of Ethidium Bromide in Laboratory 3. Wear latex

15 CAUTION Ultraviolet light can damage eyes. Never look at unshielded UV light source with naked eyes. View only through a filter or safety glasses that absorb harmful wave lengths. For further information, see Appendix Photograph with a Polaroid or digital camera. 7. Take time for responsible cleanup. a. Wipe down camera, transilluminator, and staining area. b. Wash hands before leaving lab. Destaining time is decreased by rinsing the gel in warm water, with agitation. IVB. Stain Gel with Methylene Blue and View (Photograph) (30+ minutes 1. Wear latex gloves during staining and cleanup. 2. Flood the gel with 0.025% methylene blue, and allow to stain for 20 3 minutes. 3. Following staining, use a funnel to decant as much methylene blue solutio as possible from the staining tray back into the storage container. 4. Rinse the gel in running tap water. Let the gel soak for several minutes i several changes of fresh water. DNA bands will become increasingly distinc as gel destains. For best results, continue to destain overnight in a small volume of water. (Gel may destain too much if left overnight in large volume of water.) Cover staining tray to retard evaporation. 5. View gel over light box cover the surface of the light box with plastic wra to prevent staining. 6. Photograph with a Polaroid or digital camera.

16 438 LABORATORY 8 1. A background smear of degraded and partially digested chromosomal DNA, plasmid DNA, and RNA is typically seen running much of the length of the miniprep lanes. The smear is composed of faint bands of virtually every nucleotide length that grade together. A heavy background smear, along with high-molecular-weight DNA at the top of the undigested lane, indicates that the miniprep is contaminated with large amounts of chromosomal DNA. 2. Frequently, undissolved material and high-molecular-weight DNA are seen trapped at the front edge of the well. These anomalies are not seen in commercial preparations, where plasmid DNA is separated from degraded nucleic acids by ultracentrifugation in a cesium chloride gradient. 3. A cloud of low-molecular-weight RNA is often seen in both the cut and uncut miniprep lanes at a position corresponding to 200 bp or less. Again, variously sized molecules are represented, which are the remnants of larger molecules that have been partially digested by the RNase. However, the majority of RNA is usually digested into fragments that are too small to intercalate the ethidium bromide dye or that migrate off the end of the gel. 4. Only two bands (784 bp and 3755 bp) are expected to be seen in the cut miniprep lane. However, it is common to see one or more faint bands higher up on the gels that comigrate with the uncut plasmid forms described below. Incomplete digestion is usually due to contaminants in the preparation that inhibit restriction enzyme activity or may occur when the miniprep solution contains a very high concentration of plasmid DNA. The plasmid might also be cut at only one site, creating a linear plasmid that will also migrate slower than the 3755-bp band. This is called a partial digest. It is especially confusing to see several bands in a lane known to contain only uncut plasmid DNA. This occurs because the migration of plasmid DNA in an agarose gel depends on its molecular conformation, as well as its molecular weight (base-pair size). Plasmid DNA exists in one of three major conformations: Form I, supercoiled: Although a plasmid is usually pictured as an open circle, within the E. coli cell (in vivo), the DNA strand is wound around histone-like proteins to make a compact structure. Adding these coils to the coiled DNA helix produces a supercoiled molecule. The extraction procedure strips proteins from plasmid, causing the molecule to coil about itself. Supercoiling is best demonstrated with a piece of string. Double the string and hold an end in each hand without slack. Now twist the string in one direction. At first, the coils form easily and spread evenly along the length of the string. However, as you add more twists, the string begins to bunch and form knots. If you relax the tension on the string, the string become tangled. This is the equivalent of removing the protein from supercoiled plasmid DNA. Under most gel conditions, the supercoiled plasmid DNA is the fastest-moving form. Its compact molecular shape allows it to move most easily through the agarose matrix. Therefore, the fastest-moving band of uncut plasmid is assumed to be supercoiled. Form II, relaxed or nicked circle: During DNA replication, the enzyme topoisomerase I introduces a nick into one strand of the DNA helix and rotates the strand to release the torsional strain that holds the molecule in a supercoil.

17 PURIFICATION AND IDENTIFICATION OF PLASMID DNA 439 The relaxed section of the DNA uncoils, allowing access to the replicating enzymes. Introducing nicks into supercoiled plasmid DNA produces the open circular structure with which we are familiar. Physical shearing and enzymatic cleavage during plasmid isolation introduce nicks in the supercoiled plasmid DNA. Thus, the percentage of supercoiled plasmid DNA is an indicator of the care with which the DNA is extracted from the E. coli cell. The relaxed circle is the slowest-migrating form of plasmid DNA its floppy molecular shape impedes movement through the agarose matrix. Form III, linear: Linear DNA is produced when a restriction enzyme cuts the plasmid at a single recognition site or when damage results in strand nicks directly opposite each other on the DNA helix. Under most gel conditions, linear plasmid DNA migrates at a rate intermediate between supercoiled and relaxed circle. The presence of linear DNA in a plasmid preparation is a sign of contamination with nucleases or of sloppy lab procedure (overmixing or mismeasuring SDS/NaOH and KOAc). MM294 and other strains of E. coli, termed reca +, have an enzyme system that recombines plasmids to form large concatemers of two or more plasmid units. A general mechanism for shuffling DNA strands, homologous recombination, occurs when identical regions of nucleotides are reciprocally exchanged between two DNA molecules. Homologous recombination occurs frequently between plasmids, which exist as multiple identical copies within the cell. The RecA protein binds to single-stranded regions of nicked plasmids, promoting crossover and rejoining of homologous sequences. This results in multimeric ( super ) plasmids that appear as a series of slow-migrating bands near the top of the gel. Since the concatemers form head-to-tail, they produce restriction fragments identical to those of a monomer (single plasmid) when cut with restriction enzymes. To confuse matters further, multimers can exist in any of the three forms mentioned above. Supercoiled multimers may appear further down on the gel than relaxed or linear plasmids with fewer nucleotides. 1. Examine the photograph of your stained gel (or view on a light box or overhead projector). Compare your gel with the ideal gel. Label the size of fragments in each lane of your gel. 2. Compare the two gel lanes containing miniprep DNA with the two lanes containing control pamp. Explain possible reasons for variations. 3. A plasmid preparation of pamp is composed entirely of dimeric molecules (pamp/pamp). The two molecules are joined head-to-head at a hot spot for recombination located 655 bp from the HindIII site near the origin of replication. a. Draw a map of the dimeric plasmid described above. b. Draw a map of the dimeric pamp that actually forms by head-to-tail recombination at the site described above. c. Now draw the gel-banding patterns that would result from double digestion of each of these plasmids with BamHI and HindIII, and label the basepair size of fragments in each band. 4. Explain why EDTA is an important component of TE buffer in which the miniprep DNA is dissolved.

18 440 LABORATORY 8 Component of Plasmid DNA Isolated from a reca Strain (HB101) and a reca + Strain (MM294) Ideal Gel Partial Digest

19 PURIFICATION AND IDENTIFICATION OF PLASMID DNA 441 FOR FURTHER RESEARCH 1. Isolate and characterize an unknown plasmid. Make overnight cultures of E. coli strains containing any of several commercially available plasmids (such as pamp, pkan, puc19, and pbr322). Digest miniprep and control samples of each plasmid with BamHI/HindIII, and separate by electrophoresis in an agarose gel. 2. Transform pamp and/or other plasmids into a reca + strain (MM294) and a reca strain (HB101). Do minipreps from overnight cultures of each strain, and incubate samples of each with no enzyme, HindIII, and BamHI+HindIII. Separate the samples by electrophoresis as far as possible in an agarose gel. Compare the banding patterns of the two strains, especially in the uncut lanes. 3. Research the potential use of homologous recombination in targeted gene therapy.


Why pellet and resuspend E. coli for plasmid prep - Biology

1 - Coliphage lambda DNA is a widely used vector for recombinant DNA. The middle third of its 48,000 bp contains no genes required for lytic growth and is, therefore, replaceable. The usual recombinant lambda DNA contains 80% lambda vector DNA and 20% insert, as compared to a usual cosmid DNA that contains 10% vector and 90% insert.

Wild-type lambda is not very lytic, compared to coliphages T4 or T7. Recombinant lambda phage are usually less lytic than wild-type. We recommend growing recombinant lambda phage on large (150-mm diameter) plates, rather than in liquid culture. Beginners, in particular, may not be aware sufficiently that in liquid culture E. coli can easily overgrow phage lambda, rather than the desired vice versa .

2 - Traditionally, the E. coli host for lambda is grown on NZY medium. This medium is not as rich as LB (and many other) medium, and E. coli growth is slower, making it even less likely that E. coli will overgrow the phage.

The lambda receptor on the bacterial cell outer surface is part of the maltose-usage pathway. To induce the receptor to high levels (10 2 -10 3 receptors per cell), maltose is added to a final concentration of 0.2% in the medium.

When growing lambda on plates to prepare DNA (as opposed to picking plaques or titering), agarose is substituted for agar. Agarose is much more expensive than agar, but does not contain impurities that inhibit enzymes. Lambda DNA prepared from agar plates may not be digestible by restriction enzymes because of the presence of inhibitors that have leeched from the agar.

3 - The host for lambda is E. coli C600 and its subsequent variants. To avoid complications, the E. coli C600 strain is usually restriction-negative: r-m- or r-m+ .

For our purposes, we find it useful to start with a fresh E. coli colony (on NZY or LB plates). The colony is picked into 5 ml of NZY medium plus 0.2% maltose and placed on a wheel at 37°C for approximately 4 hr. (This time will be much longer if the colony is from an old plate or has been stored at 4°C or frozen.) Using sterile conditions, the culture is transferred to a sterile 15-ml conical tube. The E. coli are pelleted by centrifugation in an SS34 rotor with adaptors in an RC-5B centrifuge (or equivalent) for 8 min at 4 K rpm (or equivalent). The medium is poured off gently. The bacterial pellet is resuspended in 4-6 ml of 0.01 M MgSO4 by pipetting up and down (using sterile conditions), NOT by vortexing. The concentration of bacteria is adjusted to yield an "optical density" (really, light scattering) of 2.0 at wavelength 600 nm. E. coli prepared in this way can be stored at room tempera ture and used for up to one week. To resuspend the E. coli , gently pipet the cells. DO NOT VORTEX. Cells prepared in this manner are employed at values of 10 µl per small or 20 µl per large plate.

4 - At times, we shall refer to the procedures published in Sambrook, Fritsch, and Maniatis, in "Molecular Cloning: a laboratory manual", pp. 2.60-2.79, second edi tion, 1991, as the "CSH" procedure or the "CSH" book.

One common source of confusion when comparing methods for growing lambda concerns "SM buffer" and whether or not it contains gelatin. In our protocol, "SM buffer" refers solely to Tris buffer with magnesium salt and NaCl. If gelatin is to be added, we state so explicitly in the few cases where we use it.

5 - The start of any procedure for preparing lambda DNA is to pick a single, well -isolated plaque. (A lambda plaque contains phage and lysed bacteria and appears relatively "clear" against the bacterial lawn.) For the Olson frozen lambda stocks, we use sterile technique to scrape a small amount (approximately a toothpick-full) of frozen stock into 1 ml of SM buffer plus gelatin. (The gelatin helps stabilize the phage.) From this one ml, we make 10-fold dilutions (in SM buffer plus gelatin) through 10-4. In general, we plate 1, 10, and 100 µl of the 10 -3 and 10-4 dilutions.

There is significant preparation time. NZY agar plates should be placed at 37°C, so that they are warm when used. NZY top agar (3 ml per plate) should be melted and placed at 50°C (just warm enough to keep it liquid, and not so hot as to kill the E. coli ). E. coli (10 µl) and 0.2 ml SM buffer (no gelatin) should be mixed in a small tube, one tube per agar plate.

Phage are added to the E. coli and SM buffer, and the tubes are placed on the 37°C wheel for 20-30 min. The kinetics of lambda attachment to its receptor are fairly slow the 20 min incubation allows time for this reaction to occur. Add the 3 ml of top agar to the mixture of phage and bacteria. Vortex for 10 sec. Pour onto an agar plate. Rock the plate by hand to reach an even distribution of top agar. Allow the top agar to solidify at room temperature (approximately 10 min). Place the inverted plates at 37°C overnight. The next day, at least one plate should have well-isolated plaques. If there are too many plaques, start over with more dilution. If there are too few plaques, start over with less dilution.

For picking plaques, we find the following procedure simple and straightforward. Autoclave inverted Pasteur pipets (without cotton plugs). Holding the narrow end, punch out a plaque with the (relatively) wide-mouth end. The agar stays in the pipet and can be expelled into a sterile eppendorf tube by a wrist flick. Add 0.5 ml of SM buffer (no gelatin) and 2 drops of chloroform (the chloroform stops the E. coli from growing). To elute the phage, allow the eppendorf tube to stand for 2-4 hr at room temperature or at 4°C overnight.

6 - A phage stock is made from the isolated plaque. Again, NZY agar plates are at 37°C NZY top agar is at 50°C 0.2 ml of SM buffer plus 10µl of E. coli are in small tubes. Add 5-20 µl (or more) of phage, depending on plaque size, to the E. coli and SM buffer. Place on a 37°C wheel for 20-30 min. Add NZY top agar and vor tex for 10 sec. Pour onto a warm NZY agar plate. After the top agar hardens (approximately 10 min), invert the plate and place at 37°C overnight.

The next morning the plates should show a fully infected, disrupted E. coli lawn. If you see individual plaques, the inoculum was too low. Toss the plates, and repeat the infection with a larger volume of the plaque. If the infection looks confluent, add 5 ml of sterile SM buffer (no gelatin). Make sure that the plate is completely covered with buffer. Gently rotate or rock the plates at room temperature for 1-2 hr. Harvest the buffer (using a sterile pipet) into a sterile conical 15-ml tube. The "milky" appearance is caused by the presence of bacterial debris. Add 0.2 ml of chloroform and vortex for 10 sec. Spin the tube in an SS34 rotor with adaptors, RC-5B centrifuge (or equivalent), at 4 K rpm for 8 min. Decant the clear yellow supernatant into a fresh, sterile, 15-ml conical tube. Add a drop of chloroform (the chloroform is to stop the growth of residual E. coli ) and store at 4°C. These stocks are stable for several months at 4°C. Frozen (-70°C) stocks can be made by mixing 0.1 ml DMSO with 1.4 ml of lambda phage stock.

7 - The phage stocks of Olson lambda are usually in the range 10 5-106 pfu (plaque-forming units) per µl. These numbers are lower than those given in CSH, pg. 2.65 the Olson lambda phage grow very poorly. Our stocks have been as low as 2x10 3 and as high as 8x107 pfu per µl. The general correlation is that phage that pro duce small plaques tend to yield stocks below 105 pfu per µl phage that produce (relatively) large plaques tend to yield stocks over 10 6 pfu per µl but the correlation is not absolute.

The lambda stocks are diluted ten-fold (as appropriate: 10 -2, 10-3, etc .) in SM buffer plus gelatin. NZY agar plates are at 37°C. NZY top agar is melted, aliquoted, and maintained at 50°C. 0.2 ml of SM buffer plus 10 µl of E. coli are in small tubes. Appropriate amounts (e.g., 1, 10, 100 µl) of each appropriate dilution are added to the bacteria. The tubes are placed on the 37°C wheel for 20-30 min. 3 ml of top agar are added to the phage and bacteria the tube is vortexed for 10 sec, and poured onto a warm agar plate. Distribute the top agar evenly on the plate while rocking (rotating, swirling) by hand. Allow the top agar to harden at room temperature (approximately 10 min). Invert the plates and place at 37°C overnight.

The next morning look at the the plaques. They are usually of a reasonable size to count. However, if the plaques are very small (which does occur with the Olson lambdas), leave the plates at 37°C and check the size of the plaques hourly. Count the plaques on plates where the number is between 50 and 300. A mini mum of 100 is needed for reasonable statistical reliability a plate with over 200 plaques may yield an incorrectly low number, as plaques overlap. How much accurracy you need is up to you. Calculate the titer of the stocks.

8 - With the preparation of titered phage stocks derived from a single plaque, all the reagents are in hard to prepare lambda DNA. In addition to titered phage stocks, you need large (150-mm) NZY agarose plates, NZY top agarose, and E. coli C600 in 0.01 M magnesium sulfate. We switch from agar to (the much more ex pensive) agarose to avoid the umwanted impurities present in agar. We usually use two 150-mm plates for growth of each original phage plaque. The large NZY agarose plates are equilibrated at 37°C. NZY top agarose is at 50°C. Place 0.2 ml of SM buffer in a small tube and add 20 µl of E. coli (double the amount for a small plate). For poor-growing lambda, such as the Olson recombinants, we add 50,000 pfu to the tube. However, this number can vary greatly: higher for very poor -growing phage and much lower for phage which grow well. You may have to test several phage inputs. The goal is to achieve confluent lysis, maximum phage pro duction, without blowing away the E. coli prematurely. Add the lambda to the bacteria and SM buffer and place the tube on the 37°C wheel for 20-30 min. Add 5 ml (an increased amount for the large plates) of top agarose vortex for 10 sec, and pour onto the warm 150-mm plate. Rotate (rock, swirl) the plate by hand to achieve an even distribution of top agarose. Allow the top agarose to harden (approximately 10 min) and invert the plate at 37°C overnight.

The next morning the plates should show confluent lysis. If individual plaques are visible, you may not achieve a usable amount of phage DNA. Start over at a higher input. We find that two confluently-lysed large plates will yield adequate amounts of DNA, even for poorly-growing lambdas.

9 - Add 8 ml of SM buffer to each plate. Make sure the buffer covers the entire plate. Sterile conditions are no longer necessary, but neatness always counts. Rotate gently (or rock gently) at room temperature for 1 to 2 hr. The phage will elute into the buffer.

Decant (pipette) the buffer into a 15-ml conical tube (or equivalent): 5-6 ml/plate. Two large plates of the same phage are combined, as appropriate. Centrifuge the tube in an SS34 rotor with adaptors, RC-5B centrifuge (or equivalent) for 8 min at 4 K rpm (or equivalent). The debris will pellet, and most of the phage will reamain in the supernatant. Decant the supernatant (9-12 ml) to an Oak Ridge tube.

10 - In this next , crucial step, DNase I is added to digest contaminating E. coli DNA. Lambda DNA, inside the phage particle, is protected from digestion. The integrity of the phage requires the presence of magnesium ions, and DNase activity requires magnesium ions, which are provided in SM buffer. Add 0.1 ml of DNase I (100 µg per ml RNase-free) and mix gently. Place the tube at 37°C for 1-2 hr. (No attempt is made to remove RNA, and, in fact, RNA is used as a carrier for the DNA.)

11 - Inactivate DNase I activity and disrupt the phage particles by the additions of 0.5 ml of 0.5 M EDTA, pH 8, and 0.5 ml of 10% SDS. Mix gently. (The removal of magnesium ions by the EDTA also causes the polysomes and ribosomes to fall apart. This helps during phage DNA purification.)

12 - To remove protein and SDS, add 10 ml of phenol (previously equilibrated with TE buffer). Vortex hard for 20 sec. Centrifuge in an SS34 rotor, RC-5B centrifuge (or equivalent), for 10 min at 10 K rpm at 10°C. Decant the upper, aqueous phase to a fresh Oak Ridge tube. (We use an inverted 10 ml pipette without cot ton plug to transfer.) Add 10 ml of chloroform. Vortex hard for 20 sec. Centri fuge in an SS34, RC-5B centrifuge (or equivalent), for 10 min at 10 K rpm at 10°C. Decant the upper, aqueous phase to a fresh Oak Ridge tube (again we use an in verted 10 ml pipette). Note the volume transferred (usually around 10 ml).

13 - Concentrate the DNA, and remove traces of phenol and chloroform, by alcohol precipitation. Add two-volumes of cold (95-100%) ethanol. Vortex for 10 sec to mix. Centrifuge in an SS34 rotor, RC-5B centrifuge (or equivalent), for at least 30 min at 12 K rpm at 10°C. Gently decant the supernatant. (We pour the alcohol off carefully while watching that the precipitate does not move.) To remove ex cess salt, gently add 10 ml of cold 70% ethanol. Do not disturb the nucleic acid precipitate. Centrifuge for 10 min (or more) at 12 K rpm at 10°C. Gently decant the supernatant. Add 10 ml of cold (95-100%) ethanol. Centrifuge for 10 min utes at 12 K rpm at 10°C. Gently decant the supernatant. Drain briefly and air dry briefly. Our experience with the Olson lambda DNAs is that the amount of precipitate is highly variable. Most of the precipitate is RNA.

14 - Dissolve the precipitate in 2.5 ml of TE. This process may take longer than 1 hr at room temperature. Add 0.1 ml RNase (1 mg/ml, DNase free) and mix gently. (At this step, the contaminating RNA is removed by digestion). Place at 37°C for 1-2 hr. Add 0.1 ml of proteinase K (1 mg/ml, nuclease-free) and incubate at 37°C for 1-2 hr. Contaminating protein, including RNase and proteinase K itself, is re moved by digestion. Cool to room temperature.

15 - Add 2.5 ml of phenol:chloroform (1:1 the phenol has previously been equili brated with TE buffer). Vortex hard for 10 sec. Centrifuge for 10 min at 10 K rpm, 10°C, in an SS34, RC-5B centrifuge (or equivalent). Decant the upper, aque ous layer to a Centricon-30 (Amicon Co.).

16 - The Centricon takes the place of alcohol precipitation and/or dialysis to change buffer and concentrate the DNA. The Centricon works annoyingly slowly, but the recoveries are excellent. Read the manufacturer's (Amicon Co.) instruc tions. The plastic that composes the Centricon is resistant to traces of phenol and chloroform but is easily damaged by isoamyl alcohol. Many recipies and commer cially available phenol:chloroform add isoamyl alcohol as an anti-foam reagent. Do not use those recipes/products, or do not use the Centricon.

We use a Centricon-30 spun at 5 K rpm in an SS34, RC-5B centrifuge at 10°C. Other rotor/centrifuge combinations may have a different maximum speed check Amicon's specifications. The choice of 10°C, rather than 4°C, is to avoid icing-up.

Under these conditions, a 2.5 ml sample will pass through the filter in about 30-60 min. We load the sample completely, do three 2.5 ml TE washes, and then spin for an additional 2 hr to reach minimum volume (usually 50-60 µl).

17 - For the Olson lambda DNAs, we use 7 µl of the DNA to double-digest and run on a 1% agarose gel in 0.5xTBE. If all has gone well, the restriction enzyme cleav age pattern matches Olson's.

Standard reagents for DNA isolation and manipulation include:

TE buffer : 0.01 M Tris, 0.001 M EDTA, pH 8.0 usually diluted from a 100x stock: 1.0 M Tris, 0.1 M EDTA, pH 8.0.

0.5 M EDTA, pH 8.0: EDTA is sold as the disodium salt. In solution, Na 2EDTA has a pH near 5 and a saturation limit around 0.2 M . To achieve 0.5 M , NaOH needs to be added in small aliquots.

10% SDS : 10 g of sodium dodecyl sulfate dissolved in water to a final volume of 100 ml.

Ethyl alcohol : 95 or 100% and 70%, both at -20°C.

Phenol : Standard commercial phenol is quite dirty by molecular biology standards and must not be used. Several companies sell re-distilled phenol of "nucleic acid" or "molecular biology" grade. We are using Amresco's saturated phenol, which is a liquid, and comes with a small bottle of TE buffer to be added to the phenol. We want buffer-saturated phenol. Add the buffer to the phenol and shake. Allow to stand overnight in a refrigerator. Use an amber bottle to avoid light. We want to avoid oxidation of the phenol to produce quinones, which are highly reactive and detrimental to DNA. A reminder: a saturated solution cannot be created in an hour. At saturation, the upper phase is excess buffer.

Chloroform (CHCl3) : any brand at A.C.S. purity is fine.

Phenol: Chloroform (1:1) : Mix one volume of buffer-saturated phenol with the same volume of CHCl3 in an amber bottle. Shake and place in the refrigerator overnight. The upper phase is excess buffer. We make our own mixture, because Amresco's phenol: chloroform contains isoamyl alcohol as an anti-foam reagent, and isoamyl alcohol dissolves the plastic Centricon.

For this procedure, the following reagents are required:

0.01 M MgSO4 : used to resuspend E. coli C600. We use A.C.S. grade MgSO4ܭH 2O. Sterilize by autoclaving or filtration.

DNase I (free of RNase) : Several companies sell this reagent. Our stock solu tion is 100 µg/ml in 0.01 M Tris, pH 8.0. We purchase DNAse I from Boehringer -Mannheim. The stock is aliquoted, frozen (-20°C), and thawed as needed. (no EDTA) Careful: DNAse I is very heat labile and is easily inactivated at room tem perature.

RNAse (free of DNAse): We purchase this enzyme from Boehringer-Mannheim. (Even the smallest amount of contaminating DNAse will degrade your precious DNA.) The RNAse is aliquoted and frozen (-20°C).

Specialized reagents for growing coliphage lambda:

  • NZ-amine (also known as "casein enzymatic hydrolysate" e.g., Sigma cat.# C -0626) 10 g
  • Yeast extract 5 g
  • NaCl 5 g
  • MgSO4ܭH2O 2 g

Sterilize by autoclaving. For agar plates (100-mm plates used for counting or isolating plaques), add 15 g of agar per liter before autoclaving. For top agar, add 7 g of agar per liter before autoclaving. For ease of handling, we prepare top agar in 100-200 ml batches, rather than 1 l batches. For large (150-mm) agarose plates, used when growing lambda to prepare DNA, add 15 g of agarose per liter before autoclaving top agarose requires 7 g of agarose per liter before autoclav ing. Unless you have an unlimited supply budget, you should use agarose ony where essential. We make both NZY-agar and NZY-agarose plates and top media to save money.

Maltose (20%) : Dissolve 20 g of maltose in a final volume of 100 ml. Sterilize by filtration. Maltose, as all sugars, tends to become caramel upon autoclaving.

(The CSH book uses the nomenclature "Tris·HCl", which is not technically correct. We start with Tris·OH and add HCl to pH 7.5. Therefore, our Tris buffers do not have additional sodium ions. Incidentally, Tris·OH is half of the cost of Tris·HCl.)


Why pellet and resuspend E. coli for plasmid prep - Biology

The biosynthesis of thiamine in Escherichia coli requires the formation of an intermediate thiazole from tyrosine, 1-deoxy- d -xylulose-5-phosphate (Dxp), and cysteine using at least six structural proteins, ThiFSGH, IscS, and ThiI. We describe for the first time the reconstitution of thiazole synthase activity using cell-free extracts and proteins derived from adenosine-treated E. coli 83-1 cells. The addition of adenosine or adenine to growing cultures of Aerobacter aerogenes, Salmonella typhimurium, and E. coli has been shown previously to relieve the repression by thiamine of its own biosynthesis and increase the expression levels of the thiamine biosynthetic enzymes. By exploiting this effect, we show that the in vitro thiazole synthase activity of cleared lysates or desalted proteins from E. coli 83-1 cells is dependent upon the addition of purified ThiGH-His complex, tyrosine (but not cysteine or 1-deoxy- d -xylulose-5-phosphate), and an as yet unidentified intermediate present in the protein fraction from these cells. The activity is strongly stimulated by the addition of S-adenosylmethionine and NADPH.


Access options

Get full journal access for 1 year

All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.

Get time limited or full article access on ReadCube.

All prices are NET prices.


Why pellet and resuspend E. coli for plasmid prep - Biology

Intestinal microbes regulate many aspects of host physiology [ 1 ], including immune system maturation [ 2 , 3 , 4 ], neurodevelopment, and behavior [ 5 , 6 , 7 ], among others. Recent reports show that in mood disorders and neurodegenerative diseases, the microbiome composition and abundance is altered, and this has provided a glimpse at the role of specific bacterial metabolites with neuroactive potential in the prevention of such disorders [ 8 , 9 , 10 , 11 ]. However, whether bacterial metabolites directly influence neuronal degeneration and their mechanisms of action are largely unknown. The bacterivore nematode Caenorhabditis elegans continues to provide an excellent model to study the relationship between bacteria and host [ 12 ]. Both the animal and its bacterial diet are genetically tractable, making them suitable for individual gene and large-scale mutation analysis. This system has been instrumental in deciphering specific metabolites from gut bacteria that influence developmental rate, fertility and aging [ 13 ], and host factors mediating germline maintenance in response to a variety of bacterial diets [ 14 ] as well as defensive behavioral strategies against pathogens [ 15 ].

Genetically encoded prodegenerative stimuli, such as a dominant gain-of-function mutation on the mechanosensory channel gene mec-4 (mec-4d) , encoding the MEC-4d degenerin, have proven effective in deciphering common molecular players of neuronal degeneration in invertebrates and mammals [ 16 , 17 , 18 ]. The touch receptor neurons (TRNs) of C . elegans respond to mechanical stimuli by causing an inward Na + current through the MEC-4 channel, a member of the degenerin/epithelial Na + (DEG/ENaC) family. Mutations near the second transmembrane (TM) helix (A713), termed mec-4d , cause the constitutive opening of the channel and the degeneration of the TRN, rendering animals insensitive to touch [ 19 ]. Necrosis of the TRNs in mec-4d worms is presumably due to unregulated Na + and Ca 2+ entry as well as reactive oxygen species (ROS) imbalance [ 18 , 20 , 21 ]. The degeneration of the TRN is a stochastic process that begins with the fragmentation of the axon followed by the swelling of the soma. Notably, the use of this model has allowed for interventions that can delay neurodegeneration, such as caloric restriction, antioxidant treatment, and mitochondria blockage [ 16 ], as well as diapause entry [ 22 ]. In this study, we evaluated the rate of degeneration of C . elegans neurons in different dietary bacteria and found that specific dietary bacteria promote protection from neuronal degeneration. Combining systems biology approaches coupled to genetics, we discovered that γ-aminobutyric acid (GABA) produced by bacteria is protective for C . elegans neurons undergoing degeneration. Moreover, further characterization indicated that GABA was not the sole metabolite involved in neuroprotection, and lactate was also identified.

Results Bacterial diet influences the rate at which neurons degenerate

We measured the effect of different dietary bacteria on the progression of genetically induced neuronal degeneration of the TRNs in a C . elegans mec-4d strain expressing a mutant mechanosensory channel, MEC-4d [ 19 ]. We previously showed that mec-4d- expressing anterior ventral microtubule (AVM) touch neuron dies in a stereotyped fashion and defined the window of time when animals feed on the standard laboratory E . coli OP50 diet [ 16 ]. Right after hatching, mec-4d mutant animals were fed different bacteria, and the AVM neuronal integrity was quantified in adulthood (72 hours later). The pertinence of the use of each of these strains is explained in the Materials and methods section. The dietary bacteria used were E . coli OP50, E . coli B, E . coli HT115, E . coli K-12, Comamonas aquatica , Comamonas testosteroni , Bacillus megaterium , and the mildly pathogenic Pseudomonas aeruginosa PAO1. As a soil nematode, C . elegans feeds on a large range of bacteria in its natural environment [ 23 ]. We also selected three bacterial species previously coisolated with wild C . elegans from soil by our group, namely Pseudochrobactrum kiredjianiae , Stenotrophomonas humi , and B . pumilus . In accordance with our previous reports, neurodegeneration steadily occurred when feeding with E . coli OP50: only a very low percentage of worms (1%–3%) maintained AVM axons after 3 days ( Fig 1A ). Notably, although neurodegeneration occurred with E . coli B, C . testosteroni , B . megaterium , and P . kiredjianiae similarly to when feeding with E . coli OP50, the bacteria E . coli HT115, E . coli K-12, C . aquatica , P . aeruginosa , S . humi , and B . pumilus gave significant protection ( Fig 1B ). E . coli HT115 was the most protective, with over 40% of wild-type axons 72 hours after hatching, compared with less than 6% in E . coli OP50 ( Fig 1B and 1D ). The broad difference on neuronal integrity in mec-4d worm populations feeding on E . coli OP50 or HT115 can be observed in Fig 1C .

10.1371/journal.pbio.3000638.g001 Fig 1 Dietary bacteria determine the rate of neuronal degeneration.

(A) TRNs expressing GFP in wild-type and mec-4d worms. The latter shows the stereotypical progression of AVM degeneration of mec-4d mutants and constitutes the axonal categories assessed during the experiment shown in (B). Scale bars represent 20 μm. (B) Percentage of all morphological axonal categories in mec-4d worms after 72 hours of growth in different bacterial strains. (C) Fluorescence microscopy fields of GFP-expressing mec-4d worms raised in the indicated E . coli strains comparing the presence of AVM axons on each preparation at 10× magnification. (D) Percentage of wild-type axons in the experiment shown in (B). (E) Percentage of touch responsiveness of animals after growth in the different dietary bacteria. **** P < 0.0001, *** P < 0.001, ** P < 0.005, * P < 0.05, ns. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. AVM, anterior ventral microtubule Axϕ, degenerated axon AxT, truncated axon AxW, wild-type axon GFP, green fluorescent protein mec-4d , mechanosensory abnormality protein 4 ns, not significant TRN, touch receptor neuron.

TRNs are neurons expressing receptors of gentle mechanical stimuli [ 24 ]. Hence, we determined the response to gentle touch in worms fed the different strains to test whether morphological protection shown in Fig 1B translates into functional responses. Fig 1E shows that the number of responses in worms correlates with the morphological categories AxW and AxL, the two axonal categories defined as functional in previous work [ 22 ].

Bacterial components promote neuroprotection

Phenotypical outcomes mediated by intestinal bacteria can be a result of either a modulation of host physiology by interspecies live interactions (i.e. bacterial colonization) or by the exposure of the host to a bacterial metabolite. The first one requires bacteria to be alive in the intestine, whereas the second does not. To distinguish between these two possibilities, we fed mec-4d animals with UV-killed HT115 bacteria, the most protective among those tested, and scored the AVM integrity at 72 hours. Additionally, we also evaluated UV-killed P . aeruginosa PAO1, a mild pathogen that needs to be alive to colonize and induce long-term defensive responses in the animal [ 15 ]. In both cases, dead bacteria protected to the same extent as live bacteria ( Fig 2A ). This indicates that protective molecules of E . coli HT115 bacteria are produced prior to exposure to the animals, and thus neuronal protection is independent of the induction of bacterial responses by the interaction with the host. Furthermore, worms raised in dead HT115 bacteria cultivated to different optical densities (ODs) displayed the same levels of neuroprotection, suggesting that the protective factors are present in the bacteria during all phases of the growth curve ( S1A and S1B Fig ).

10.1371/journal.pbio.3000638.g002 Fig 2 Bacterial components have neuroprotective activity.

(A) Axonal categories in worms raised in the indicated live or UV-killed bacterial strains. (B–C) All axonal categories (B) or wild-type (C) axons in worms rose in different proportions of UV-killed E . coli HT115 on UV-killed OP50. (D) Wild-type axons of mec-4d animals fed with E . coli HT115 or OP50 bacterial pellet supplemented with supernatant of OP50 or HT115. **** P < 0.0001, *** P < 0.001, ** P < 0.005, * P <0.05 ns. “a” and “b” are used to indicate statistically significant differences. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. Axϕ, degenerated axon AxT, truncated axon AxW, wild-type axon LB, Luria-Bertani mec-4d , mechanosensory abnormality protein 4 NGM, nematode growth media ns, not significant.

The large difference in neuroprotection between E . coli OP50 and HT115 strains raises two possibilities: (1) E . coli OP50 actively promotes the degeneration of the neuron, and (2) HT115 has a protective effect. To discern this matter, we fed worms with a mix of UV-killed E . coli HT115 and OP50 in different proportions ( Fig 2B and 2C ). A 1/100 (1%) dilution of E . coli HT115 in OP50 was sufficient to protect AVM neurons significantly more than pure OP50. This strongly suggested that E . coli HT115 produces a neuroprotective compound that is needed in small amounts.

To test whether the neuroprotective molecules are being secreted by the bacteria, we separated the supernatant of both bacterial strains from their pellets by centrifugation and mixed the supernatant of E . coli HT115 with OP50 pellet and vice versa (see Materials and methods for details). E . coli HT115 supernatant was not capable of providing protective activity when mixed with E . coli OP50 pellet ( Fig 2D ). This suggests either that the protective factor is not secreted or that the amounts contained in the supernatant are not sufficient for protection. As expected, E . coli OP50 supernatant did not alter the protection pattern of E . coli HT115.

E . coli HT115 diet promotes long-term protection of mechanoreceptors and interneurons of the touch receptor circuit

E . coli HT115 was shown to be neuroprotective throughout the development of the animal and into young adulthood ( Fig 1B ). We explored whether AVM neurons are still protected after worms reached maturity. To that end, we fed newly hatched mec-4d animals with E . coli HT115 and scored their neuronal integrity every 24 hours for 168 hours. While on E . coli OP50, all animals had degenerated neurons at the final time point, and on HT115 food, 25% of animals had wild-type AVM axons ( Fig 3A and 3C ), confirming the ability of HT115 to significantly protect at later life stages. Notably, between 12 and 24 hours after hatching in HT115, there was a statistically significant rise in wild-type axons (AxW, Fig 3C ), suggesting that neurons could be growing after an initial truncation. To assess this, we followed individual animals in a longitudinal fashion on E . coli HT115 and scored the neuronal integrity of each nematode every 24 hours for 3 days. We scored axons separately according to their initial and final morphology and classified axonal outcome as “protection” when the morphology of axons did not change from truncated or wild type and “degeneration” when axonal morphology changed from truncated axon (AxT) to degenerated axon (Axϕ) or was maintained as Axϕ. Finally, “regeneration” refers to axon growth from truncations to wild type. Although the most prevalent category is protection (40%), 30% of axons regenerated between 24 and 72 hours after hatching on E . coli HT115 ( Fig 3D ) . This suggests that under HT115 protective conditions, a portion of neurons can repair broken axons.

10.1371/journal.pbio.3000638.g003 Fig 3 Neuroprotection induced by dietary E . coli HT115 is long lasting and extends to other neuronal types.

Time course of axonal categories of worms fed with E . coli OP50 (A) or HT115 (B) for 168 hours. (C) Percentage of wild-type axons in (A) and (B). (D) Proportion of axonal categories in longitudinal assays. Protection (green) indicates axons that were not degenerated over time regardless of the initial category, with the exception of Axϕ. Regeneration (gray) accounts for axons that grew in size over time. Degeneration (red) accounts for axons that degenerated over time. Determination of wild-type axons in AVM (E), ALM (F), PLM (G), and PVC (H) neurons of worms fed E . coli OP50 or HT115. **** P < 0.0001, *** P < 0.001, ns. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. ALM, anterior lateral microtubule AVM, anterior ventral microtubule Axϕ, degenerated axon AxT, truncated axon AxW, wild-type axon ns, not significant PLM, posterior lateral microtubule PVC, ventral cord interneuron PVM, posterior ventral microtubule.

Next, we explored whether other neurons of the touch circuit are protected from degeneration in E . coli HT115 diet. It has been already reported that at hatching, four embryonic TRN, two anterior lateral microtubule (ALMs), and two posterior lateral microtubule (PLM) neurons have already degenerated when growing at 20 °C in this model [ 16 , 25 ]. At 25 °C, however, degeneration proceeds at a slower rate [ 16 , 22 ]. To analyze the degeneration rate of ALM and PLM neurons, animals in the fourth larval stage (L4) were grown at 25 °C, and their progenies were synchronized at birth. The neuronal integrity of ALM, PLM, and AVM cells was assessed at 12, 24, and every 24 hours after birth until 168 hours at 25 °C. The percentage of wild-type neurons in E . coli HT115 diet is significantly higher than that in the OP50 diet throughout the temporal course for all three neurons ( Fig 3E and 3G , full morphological characterization is shown in S2A and S2F Fig ).

We tested next whether E . coli HT115 was capable of protecting the ventral cord interneuron (PVC) expressing the degenerin-1 ( deg-1 ) prodegenerative stimulus. deg-1(u38) animals progressively lose the ability to respond to posterior touch due to the time-dependent degeneration of the PVC interneuron [ 26 ]. We tested the posterior touch response of deg-1 animals during development feeding on E . coli OP50 and HT115. E . coli HT115 promotes a larger functional response than E . coli OP50, suggesting that this neuron is also protected ( Fig 3H ). Taken together, these results demonstrate that HT115 diet is protective over different neuronal types undergoing degeneration.

Neurodegeneration of the TRNs is directly related to the expression of a neurotoxic form of the mechanosensory channel. Therefore, it is formally possible that a decrease in the expression of the channel would diminish the prodegenerative stimulus and promote protection. We sought to evaluate whether HT115 diet changes the expression of the MEC-4d channel in the membrane. We constructed a double mutant of mec-4d expressing MEC-4::green fluorescent protein (GFP) and quantified the number of channels of PLMs in HT115-fed animals compared with OP50. S3A and S3B Fig show that channel number remains constant in both diets, ruling out that protection conferred by HT115 affects expression of MEC-4d channel in the membrane.

Early exposure of animals to E . coli HT115 is sufficient for neuronal protection

Ad libitum feeding on E . coli HT115 protected mec-4d -expressing neurons from degeneration for long periods of time. We sought to investigate whether a constant stimulus provided by the HT115 metabolite is required to achieve neuroprotection or an early, discrete time-lapse exposure to the diet is sufficient. We fed animals for the first 6 hours after hatching (previous to the birth of the AVM) and for 12 hours after hatching (at birth of the neuron) with UV-killed E . coli HT115 and immediately switched to E . coli OP50. We scored the neuronal morphology 12, 24, 48, and 72 hours posthatching ( S4A and S4D Fig ). In parallel, both diets were fed ad libitum as controls. Strikingly, animals that ingested E . coli HT115 for only 6 hours showed a significantly larger number of wild-type neurons at 72 hours (14.3%) than animals continuously fed OP50 (3.6%, Fig 4A ) and had more axons in the other categories ( S4A and S4C Fig ). Feeding HT115 to mec-4d animals for the first 12 hours after hatching conferred a significantly larger protection than feeding HT115 for only 6 hours ( Fig 4A ). These results show that although early short exposures are not equally protective as a permanent HT115 diet, they do have a long-lasting effect in neurons compared with an uninterrupted diet of E . coli OP50.

10.1371/journal.pbio.3000638.g004 Fig 4 Diet of E . coli HT115 at early stages of life is necessary and sufficient to confer neuroprotection.

(A and B) Percentage of wild-type axons of animals fed for 6 and 12 hours with (A) E . coli HT115 or (B) E . coli OP50 and then changed to OP50 or HT115 diet, respectively. (C) Percentage of wild-type axons of animals feeding on either E . coli OP50 or HT115 whose parents were fed on either diet. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. F, filial generation P, parental generation.

We then tested the effect of early exposure to nonprotective bacteria. We fed mec-4d animals for 6 and 12 hours with UV-killed E . coli OP50 and then changed them to HT115. The morphology of AVM neurons was scored at 12, 24, 48, and 72 hours posthatching. Six hours of E . coli OP50 exposure did not prevent HT115 from protecting AVM neurons later in adulthood ( Fig 4B and S4E Fig ). Exposure for 12 hours, however, precludes protection of the AVM ( Fig 4B and S4F Fig ). This suggests that the time between the first 6 and 12 hours of development is crucial for the protective effect to take place.

In C . elegans , some dietary bacteria–induced traits show heritable properties [ 15 ]. Therefore, we tested whether neuronal protection could be inherited. Animals were fed either E . coli OP50 or HT115 from birth, and their F1 progeny transferred to OP50 or HT115. Neuronal integrity of descendants was tested in a time course fashion. One generation of parental feeding on HT115 did not improve neuronal protection in the progeny feeding on OP50, nor did E . coli food preclude protection of filial generation (F) 1 feeding on HT115 ( Fig 4C ). This result indicates that the protective effect of E . coli HT115 is not transmitted intergenerationally.

Identification of uniquely expressed genes on neuroprotective bacteria

To identify the bacterial molecule(s) conferring neuroprotection, we looked for differences in the genomes and transcriptomes of the two E . coli strains. We reasoned that genes important for neuronal protection would be uniquely expressed or up-regulated in E . coli HT115 compared with E . coli OP50. We first sequenced the genomes of E . coli HT115 and OP50 using the Illumina MiSeq platform. Notably, we found that E . coli OP50 has a deletion of 23 Kbp, containing genes for the regulators of capsular system ( rcsDB , rcsC ) required for envelope stress response, genes for short-chain polyhydroxybutyrate synthesis ( atoSC , atoDAEB ), and the pseudogene yfaATSQP [ 27 ]. Thus, RcsB is a unique E . coli HT115 gene that codes a main transcription factor required for the activation of the glutamate decarboxylase enzyme gene ( gad ) operons, alone or coupled to other regulators by positive feedback as illustrated in Fig 5A .

10.1371/journal.pbio.3000638.g005 Fig 5 Enzymes of the GAD enzyme operon are uniquely and highly expressed in E . coli HT115.

(A) Scheme of regulation of the gad operon. (B) Log2 Fold Change of genes coding for enzymes of GABA metabolism. The underlying numerical data and statistical analysis for each figure panel can be found in S1 Dataset and S2 File , respectively. GABA, γ-aminobutyric acid GAD, glutamate decarboxylase gad , glutamate decarboxylase enzyme gene rcs , regulator of capsular system.

Transcriptomic analysis determined that the genes involved in resistance to acidic environments were highly up-regulated in E . coli HT115 ( Fig 5B ). The glutamate decarboxylase operons gadAX and gadBC include the genes gadA and gadB , both encoding for the enzyme glutamate decarboxylase (GAD), which converts glutamate to GABA, whereas gadC encodes a glutamate/GABA antiporter. Other overexpressed genes in HT115 include those for the periplasmic acid stress chaperones hdeA and hdeB . Importantly, this overexpression occurs while global regulators also involved in the acidic response remain equally expressed in both strains (CRP-AMPc, H-NS, or Fis). Therefore, rcsB deletion seems to induce a specific metabolic difference between both bacteria, i.e., the abolition of GABA production in E . coli OP50. To endure this deficit, genes related to sodium/glutamate and glutamate/aspartate transport were up-regulated in E . coli OP50 as shown in Fig 5B ( gltS Log 2 Fold Change (LgFC) = −4.69, gltK LgFC = −2.08, respectively). Additionally, other genes related to GABA metabolism (the transaminase gabT LgFC = 2.67) and membrane permeability (permease gabP LgFC = 2.74) were also up-regulated in E . coli HT115 from non- rcsB -dependent operons. Interestingly, no other metabolic enzyme-related genes were up-regulated in either bacterium. This suggests that enzymes and metabolites involved in the pathway of GABA production and utilization are good candidate neuroprotective players.

GAD and its product GABA are required for E . coli HT115 neuroprotection

To test the role of GAD and its product GABA in neuroprotection, we first generated a gad null mutant of E . coli HT115 by homologous recombination (HT115Δ gad , details in Materials and methods ). To corroborate that HT115Δ gad lacked GAD activity, we used a colorimetric assay based on pH elevation given by the conversion of glutamate to GABA [ 28 , 29 ]. As expected, wild-type E . coli HT115 raised the pH of the solution, whereas neither HT115Δ gad nor OP50 were able to do so. To confirm that a rise in pH is due to the expression of GAD, we transformed E . coli OP50 with a plasmid expressing gadA (pG gadA ). E . coli OP50 pG gadA supplemented with glutamate showed potent enzymatic activity, raising the pH of the solution above HT115 levels ( Fig 6A ). Secondly, we fed mec-4d animals with E . coli HT115Δ gad and scored its protective potential at 72 hours. HT115Δ gad was not able to protect degenerating AVM neurons, showing a significant reduction of wild-type axons compared with wild-type strains ( Fig 6B ). This shows that GAD activity plays a pivotal role in the protection conferred by HT115 bacteria. Moreover, plasmid pG gadA was able to rescue protective potential in null mutant HT115Δ gad . Additionally, dietary supplementation of UV-killed HT115Δ gad with 2 mM of GABA was sufficient to provide neuroprotection ( Fig 6B and S5A Fig ).

10.1371/journal.pbio.3000638.g006 Fig 6 Bacterial GABA is crucial for neuroprotection.

(A) Measurements of GAD enzyme activity normalized as a percentage of HT115 GAD activity in wild-type, mutant, and transformed bacterial strains. (B and C) Percentage of wild-type axons in wild-type and Δ gad mutant HT115 strain (B) and wild-type OP50 strain (C) supplemented with GABA and genetically transformed with the pG gadA plasmid. **** P < 0.0001, *** P < 0.001, ns. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. GABA, γ-aminobutyric acid GAD, glutamate decarboxylase gad , glutamate decarboxylase enzyme gene ns, not significant.

Finally, we fed mec-4d with E . coli OP50 pG gadA to test whether gadA is sufficient to provide protective activity in the presence or absence of glutamic acid, the substrate for GAD. Whereas E . coli OP50 pG gadA alone was not sufficient to increase wild-type axons incidence, glutamate addition to the bacterial culture significantly increased the presence of wild-type axons compared with E . coli OP50 pG gadA and E . coli OP50 wild type ( Fig 6C and S5B Fig ). This is coherent with the increased GAD activity of E . coli OP50 supplemented with pG gadA shown in Fig 6A . Furthermore, supplementation of HT115Δ gad with 2 mM GABA was sufficient to provide neuroprotection ( Fig 6C ). Importantly, addition of 2 mM GABA to UV-killed E . coli OP50 lawn protected mec-4d neurons significantly more than OP50 alone, even though it did not reach HT115 levels ( Fig 6C ). Taken together, these results show that GAD and its product GABA play an important role in E . coli HT115–mediated neuroprotection.

Identification of metabolites expressed in neuroprotective conditions

To unbiasedly identify potentially neuroprotective metabolites produced by the strain HT115 but absent in the nonprotective HT115 Δ gad and E . coli OP50 strains, we implemented a nontargeted metabonomics approach using 1H nuclear magnetic resonance (NMR). A total of 24 extract samples were prepared and analyzed (eight of each strain). To evaluate the global metabolic profile of the three bacterial strains, we performed a principal component analysis (PCA) of binned 1H NMR datasets. As we expected, all three bacteria strains were metabolically different, with E . coli HT115 and HT115 Δ gad closer than E . coli OP50 in the metabolic space ( S5A Fig ). Metabolites related with neuroprotection were evaluated by orthogonal projections to latent structures discriminant analysis (OPLS-DA), first comparing E . coli HT115 strain against HT115 Δ gad ( Fig 7A ), and secondly, comparing E . coli HT115 against HT115 Δ gad and OP50 ( Fig 7D ). OPLS-DA models were validated by 200 permutations ( S6B and S6C Fig ). Discriminant analysis in HT115 versus HT115 Δ gad revealed intergroup metabolic differences. The discriminant up-regulated metabolites in wild-type HT115 were GABA, lactate, sucrose, and maltose, whereas in HT115 Δ gad , they were glutamate and putrescine ( Fig 7B and 7C , S1 Table and S7 Fig ). This is coherent with the accumulation of the GAD substrate glutamate, given the absence of the enzyme on HT115 Δ gad . Notably, the comparison between E . coli HT115 strain against HT115 Δ gad and E . coli OP50 revealed the following intergroup metabolic differences: GABA, lactate, sucrose and maltose were highly expressed in the neuroprotective strain, whereas there were no discriminant metabolites found in higher levels in OP50 and HT115 Δ gad ( Fig 7E and 7F and S2 Table ). Overall, in strong agreement with our genetic and chemical complementation approach, these results further indicated that GABA is one of the metabolites playing a central role in HT115-conferred neuroprotection.

10.1371/journal.pbio.3000638.g007 Fig 7 Metabolomics analysis of neuroprotective and nonprotective bacteria.

(A–D) OPLS-DA score plot of protective E . coli HT115 wild type (blue) and nonprotective E . coli HT115Δ gad (light gray) (A) and E . coli HT115 wild type (blue) and nonprotective E . coli HT115Δ gad (light gray) and OP50 (dark gray) (D). (B–E) OPLS-DA S-line plots with pairwise comparison of data from NMR spectra obtained comparing the E . coli HT115 strain against HT115 Δ gad (B) and the E . coli HT115 strain against HT115 Δ gad and E . coli OP50 (E). Colors are associated with the correlation of metabolites characterizing the 1H NMR data for the class of interest, with the scale shown on the right-hand side of the plot. In (B), GABA and glutamate signals are shown. (C–F) Tables indicate which metabolites are differentially expressed in each strain. (G) Percentage of wild-type axons in mec-4d animals fed with E . coli OP50 supplemented with lactate. *** P < 0.001 ns. The underlying numerical data for the figure panels can be found in S1 and S2 Tables (A to F) and S1 Dataset (G). Statistical analysis can be found in S2 Dataset . GABA, γ-aminobutyric acid gad , glutamate decarboxylase enzyme gene mec-4d , mechanosensory abnormality protein 4 NMR, nuclear magnetic resonance ns, not significant OPLS-DA, orthogonal projections to latent structures discriminant analysis.

Importantly, the supplementation of E . coli OP50 with GABA did not reach HT115 neuroprotective levels. We wonder whether other metabolites overabundant in E . coli HT115 could also contribute to neuronal protection. Lactate is a metabolite not expressed either in the Δ gad or OP50 bacteria. To test the role of lactate in neuroprotection of C . elegans TRNs, we fed animals with E . coli OP50 bacteria supplemented with 2 mM lactate. Fig 7G and S8 Fig show that lactate confers significant protection to E . coli OP50 bacteria. This suggests that lactate in addition to GABA contribute to neuroprotection.

GAD activity and GABA levels correlate with neuroprotection conferred by other bacteria

The amount of a neuroprotective metabolite present in a given bacterium could be an indicator of that bacteria’s ability to confer neuronal protection. Bacteria tested by us in Fig 1B gave different degrees of protection to mec-4d animals. We evaluated GAD activity in all strains and normalized it against E . coli HT115 ( Fig 8A ). All strains except E . coli K-12 exhibited less GAD activity than E . coli HT115. Additionally, we directly measured GABA production using the GABA‐aminotransferase plus succinic semialdehyde dehydrogenase (GABase test [ 30 ] and S9 Fig ). E . coli HT115 pellet had the highest GABA levels, whereas OP50 and HT115Δ gad were indistinguishable from each other ( Fig 8B ). This demonstrates that GABA is being produced in E . coli HT115 and not in OP50 or the HT115 Δ gad strain. P . aeruginosa PAO1 and B . pumilus had less GABA than HT115 but significantly more than most strains ( Fig 8B ). To understand whether there was a correlation between GAD and GABA levels with neuroprotection conferred by these bacteria, we performed a Pearson correlation test. Fig 8C shows that GAD expression and GABA levels are correlated with neuroprotective activity in all strains. Importantly, GABA concentration was a better indicator (r = 0.88) than GAD activity (r = 0.67). These results support the previous evidence that bacterial GAD enzyme and its product GABA are key for neuroprotection.

10.1371/journal.pbio.3000638.g008 Fig 8 GAD activity and GABA levels correlate with neuroprotection conferred by other bacteria.

Measurements of GAD enzyme activity normalized as a percentage of HT115 GAD activity (A) and GABA concentration found in pellets (B) in all bacteria used. (C) Correlation between GAD activity and GABA concentration in bacteria diet and percentage of wild-type axons in C . elegans . **** P < 0.0001, ** P < 0.005 ns. “a,” “b,” “c,” and “d” are used to indicate statistically significant differences. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. GABA, γ-aminobutyric acid GAD, glutamate decarboxylase ns, not significant.

Host GABA receptors and transporters are required for full HT115 bacteria–mediated neuroprotection

To discern whether systemic or neuronal GABA transport was implicated in neuroprotection, we silenced the expression of a number of candidate solute transporters ( unc-47 , snf-5 ) and GABA receptors ( gab-1 , lgc-37 , unc-49 ) using RNA interference (RNAi). To distinguish a systemic from a touch cell–specific requirement for these effectors, we used mec-4d animals (in which neuronal RNAi is inefficient) and mec-4d animals sensitive to RNAi only in the TRN (WCH6, [ 16 ]). We fed double-stranded RNA (dsRNA) of the selected effectors to both strains, and the neuronal morphology was assessed at 72 hours. lgc-37 , snf-5 , unc-47 , and gab-1 dsRNA-expressing bacteria caused a discrete but significant decrease in neuronal protection in the systemic RNAi strain but not in the TRN-specific strain ( Fig 9A and 9B and S10A and S10B Fig ), suggesting these genes act in nonneuronal tissues to mediate neuroprotection.

10.1371/journal.pbio.3000638.g009 Fig 9 Effect of silencing GABA effectors in C . elegans on neuroprotection.

Morphological integrity of AVM in animals feeding on dsRNA-expressing bacteria for the indicated genes in a systemic (A) and TRN-autonomous (B) RNAi strain. ** P < 0.005, * P < 0.05 ns. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. AVM, anterior ventral microtubule dsRNA, double-stranded RNA GABA, γ-aminobutyric acid gad , glutamate decarboxylase enzyme gene ns, not significant RNAi, RNA interference TRN, touch receptor neuron.

E . coli HT115 protection requires DAF-16 signaling

We explored the role of the insulin/IGF-1-like signaling (IIS) pathway, a well-described and conserved signaling cascade acting systemically. In C . elegans , down-regulation of the insulin receptor DAF-2 is neuroprotective [ 16 ]. We investigated whether the E . coli HT115 neuroprotective effect also involved the IIS. First, we fed daf-2ts mec-4d animals with E . coli HT115 at 25 °C. This strain expresses a DAF-2 protein version that is unstable at 25 °C. Next, we scored the neuronal integrity of ALM, PLM, and AVM neurons in a time-dependent fashion. ALM and PLM protection in E . coli HT115 was not further increased by the down-regulation of DAF-2 ( Fig 10A and 10B , neuronal integrity on E . coli OP50 of ALM, PLM, and AVM at 25 °C in S11A and S11C Fig ), suggesting that HT115-mediated protection involves the down-regulation of the DAF-2 pathway. Owing to the cumulative protection effects of temperature and diet, mec-4d AVM neurons at 25 °C on HT115 diet reached almost 100% of wild-type axons, and the daf-2 mutation maintained maximum protection ( Fig 10C ).

10.1371/journal.pbio.3000638.g010 Fig 10 Effect of down-regulation of the insulin pathway on E . coli HT115–mediated neuroprotection.

(A–C) Neuronal integrity of ALM (A), PLM (B), and AVM (C) neurons of daf-2(ts ) mec-4d animals feeding on HT115 food. (D) Number of GFP-positive nuclei in DAF-16::GFP animals feeding on E . coli OP50 or HT115 during development. DAF-16::GFP is observed in the entire body. (E) Degree of GFP expression of animals in (D). (F and G) Time course of neuronal degeneration in daf-16 mec-4d animals at 20 °C (E) and 25 °C (F). **** P < 0.0001, ** P < 0.005 ns. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. ALM, anterior lateral microtubule AVM, anterior ventral microtubule DAF-2, ortholog of the insulin receptor DAF-16, ortholog of the FOXO transcriptional factor GABA, γ-aminobutyric acid gad , glutamate decarboxylase enzyme gene GFP, green fluorescent protein mec-4d , mechanosensory abnormality protein 4 ns, not significant PLM, posterior lateral microtubule.

Because DAF-2 down-regulation causes the translocation of DAF-16 (ortholog of the FOXO family of transcription factors) to nuclei [ 31 ], we tested whether HT115 diet promoted nuclear translocation of GFP in a DAF-16::GFP-expressing strain (CF1139 strain, see Materials and methods ). We fed CF1139 animals with E . coli OP50 and HT115 and compared the number of fluorescent nuclei in the entire body of animals at 12, 24, and 48 hours after hatching on each diet. Additionally, we qualified the intensity of GFP expression (see Materials and methods ). E . coli HT115 promoted a significantly higher translocation of DAF-16 compared with E . coli OP50 at 24 hours only, returning to basal levels at 48 hours ( Fig 10D ). Given that bacterial GAD enzyme and GABA are correlated with the protection conferred by HT115 ( Fig 8C ), we tested whether they were involved in DAF-16 nuclear translocation. The Δ gad mutation caused a significant decrease in both the number of GFP-positive nuclei and the intensity of GFP expression ( Fig 10D and 10E ). This decrease, however, did not entirely eliminate DAF-16 nuclear expression, suggesting that in the context of E . coli HT115, GAD enzyme is not the only factor responsible for activation of DAF-16. The addition of GABA to E . coli OP50, although effective in conferring neuroprotection, failed to cause nuclear translocation of DAF-16. Similarly, lactate supplementation of E . coli OP50, though strongly protective of C . elegans AVM axons, also failed to promote DAF-16 translocation. This suggests that protection by bacterial metabolites GABA and lactate is likely caused by a mechanism independent of further translocation of DAF-16 to nuclei.

Next, we directly assessed the involvement of DAF-16 in the neuroprotective effect of E . coli HT115 by scoring neuronal integrity of the TRNs in daf-16 mec-4d animals feeding with HT115 at 20 °C and 25 °C. Both in E . coli OP50 and HT115, most TRN types were absent at birth in either temperature, with a marginal presence of AVMs, which also rapidly underwent degeneration in the daf-16 mutant ( Fig 10F and 10G , all categories in E . coli OP50 and HT115 at 20 °C and 25 °C in S11D and S11G Fig ). This demonstrated that DAF-16 is required for the neuroprotection effect of the HT115 diet. Notably, the daf-16 mutation completely abolished the protection of TRNs previously observed at 25 °C in mec-4d background [ 22 ] in both bacteria, suggesting that the daf-16 mutation significantly lowers the threshold for neurodegenerative stimuli. The effect of increased degeneration in daf-16 mec-4d , though much more dramatic in E . coli HT115, is also observed in OP50. We assessed the effect of daf-16 mutations in normal development and integrity of the TRNs in the strain WCH40 ( daf-16 [ m27 ] uIs31 [ P mec-17 mec-17 :: gfp ]) expressing GFP in all the TRNs. DAF-16 loss alone did not cause an observable effect in the morphology of the TRNs ( S11H Fig ). Taken together, these results show that the function of DAF-16 is required for E . coli HT115–mediated protection to take place.

The relationship between bacteria and host affects virtually every studied aspect of an animal’s physiology. However, whether bacteria and their metabolites can modulate neuronal degeneration is not known. In this work we show that bacterial diet dramatically influences neuronal outcomes in a C . elegans model of neurotoxic death triggered by the MEC-4d degenerin. E . coli HT115, a derivative of the K-12 strain, is the most protective of all bacteria tested, which included nonpathogenic laboratory bacteria, mild pathogens, and natural commensal bacteria. We found that this bacterium protects embryonic and postembryonic TRNs as well as the PVC interneuron, suggesting a pleiotropic effect on the nervous system. By comparing the genomes, transcriptomes, and metabolomes of the most and least protective E . coli strains, we found that GABA is a key bacterial neuroprotective metabolite. Systemic C . elegans GABA receptors GAB-1 and LGC-37 and GABA transporter UNC-47 are required for wild-type neuroprotection conferred by HT115 bacteria. In addition to GABA, bacterial lactate was able to confer large neuroprotection. Importantly, HT115 neuroprotective effect can only take place on the condition of a functional DAF-16 transcription factor.

Gut microbes regulate neurodegeneration in C . elegans

Recent work has highlighted the importance of the gut microbiota in shaping human health and well-being. Not only do intestinal microbes regulate many aspects of host physiology and development, but they have also been linked to mood disorders and neurodegenerative diseases [ 7 , 32 , 33 ]. For example, patients with Parkinson disease (PD) present constipation as a nonmotor symptom in early stages of the disease, which correlates with dysbiosis of the gut microbiome. Although it is unclear whether alterations in the gut microbiome are causes or consequences of these illnesses in the nervous system, emerging evidence using fecal transplantation in animal models demonstrates the ability of healthy and diseased gut microbiomes to ameliorate symptoms and confer disease, respectively [ 34 , 35 ].

Intestinal microbiota regulates central nervous system (CNS)-related traits through the microbiota–gut–brain axis. This consists of a bidirectional communication between the microbiota in the intestinal tract and the brain through the production of neuroactive molecules. Few recent studies correlate bacterial taxa from mammal’s microbiota with metabolite production and its impact on brain function and pathology [ 36 ].

The bacterivore nematode C . elegans offers an advantaged platform for dissecting specific neuroprotective metabolites because bacteria can be given monoaxenically to animals that express a genetically encoded degenerative trigger ( mec-4d ). Dietary bacteria are usually broken up in the grinder located in the worm’s pharynx right after ingestion [ 37 ]. This releases bacteria contents into the gut of the worm, a process that may also occur by explosive lysis [ 38 , 39 ] Nonetheless, a number of dietary bacteria survive this initial interaction to live and colonize the worm’s digestive tract [ 40 ]. We show that UV-killed E . coli HT115 is equally as neuroprotective as live bacteria, ruling out that bacteria need to colonize the intestine and actively trigger a host immunoprotective response. Moreover, HT115 fed only for a few initial hours to newly hatched animals was sufficient to provide protection to adult neurons. This suggests that HT115 metabolites turn on a signaling cascade that outlasts the presence of the protective bacterial metabolite itself.

Bacterial GABA is neuroprotective

The GAD system is a key mechanism used by intestinal bacteria to cope with acidic stress. This enzyme that produces GABA from glutamic acid is uniquely expressed in E . coli HT115. Genes coding for GAD are conserved in the genome of E . coli OP50 nonetheless, this strain lacks rcsB , encoding a transcriptional factor known to be required for the expression of gad genes in E . coli . A close inspection to the genome of OP50 reveals that this strain lost a chromosomal segment spanning from micF up to yfaA , which includes rcsBCD genes. This region is substituted by an IS1 in OP50, which could account for the low production of GABA in this strain. Notably, it has been reported that growth of an E . coli K-12 strain in buffered acidic conditions promotes the loss of several systems employed to grow in unbuffered acidic conditions, including GABA production [ 41 ]. OP50 is a derivative of E . coli B2. Like OP50, the B2 strain has a diminished GABA production and also lacks the rcs region. Thus, this and other traits may resemble the different selective pressures that have been present over the E . coli K-12 and B lineages. Nonetheless, HT115 has a highly increased capacity to produce GABA, even compared with other K-12 strains. In fact, even among different unrelated bacteria tested, only P . aeruginosa PAO1 seemed to produce similar levels of GABA ( Fig 8B ). The origin of GABA overproduction in HT115 is not evident from genomic analyses and requires further research.

GABA can be metabolized to succinic semialdehyde (SSA) and succinate or translocated to the periplasm by the glutamate/GABA antiporter (GADT, [ 42 ]). GADT is dramatically overexpressed in protective bacteria, likely tilting the balance toward the accumulation of GABA in the periplasm. Unbiased metabolomic analysis identified GABA as one of the main metabolites differentially expressed in E . coli HT115 compared with OP50. Finally, GABA supplementation to E . coli OP50 is sufficient to provide neuroprotection.

Some human enteric microbiota members have been shown to require GABA to grow, and thus the production of GABA by enteric members of the Bacteroides , Parabacteroides , and Escherichia genera has been suggested to delineate the composition of the human microbiota [ 9 ]. Moreover, in the same work, it was shown that the relative abundance of GABA-producing members in intestinal microbiota negatively correlates with depression-associated brain signatures in patients. Bacterial GABA was proposed as one of the main effectors of microbiota on the CNS [ 8 , 9 ]. How bacterially produced GABA in the gastrointestinal tract may affect the brain or other distal CNS traits is intriguing. In animals, GABA receptors are expressed in epithelial cells, and a limited capacity of GABA to cross the blood–brain barrier has been reported [ 43 ]. However, it is not clear whether GABA and other microbiota-produced neuroactive molecules may exert their effect by localized or systemic activation of signaling pathways. In C . elegans , the number and type of GABA-containing cells and cells expressing GABA-uptake proteins and receptors are higher than previously thought and include nonneuronal cells [ 44 ]. Thus, although no intestinal GABAergic cells were reported, GABA could be pleiotropically sensed by a number of cells. In an infection model, Staphylococcus aureus molecules can trigger neuroendocrine reactions in C . elegans [ 45 ]. In our work, by using systemic and TRN-specific reverse genetics, we show that systemically delivered dsRNA of unc-47 , lgc-37 , and gab-1 decreases neuroprotection in HT115 bacteria. This shows that GABA can be processed in nonneuronal cells to promote neuroprotection.

DAF-16/Forkhead box O transcription factor loss impairs E . coli HT115 protection

A phosphorylation cascade downstream of the insulin receptor DAF-2/IGF1R controls DAF-16 transcription factor activation. When DAF-2/IGF1R is activated, DAF-16/Forkhead box O (FOXO) is prevented from entering the nuclei [ 31 ]. Starvation, pathogen exposure, and other interventions promote its nuclear translocation and activation of transcriptional targets [ 46 ]. DAF-16 can also be directly activated upon fungal infection in a DAF-2-independent fashion [ 47 ]. In our experiments, DAF-2 inactivation did not increase HT115 protection, suggesting they act in the same pathway. Neuroprotective and neuroregenerative effects of DAF-2 down-regulation involve the function of DAF-16 [ 16 , 22 , 48 ]. Comparison of DAF-16 nuclear translocation between diets showed that at 24 hours DAF-16 was increased in the nuclei of animals feeding on E . coli HT115 compared with OP50. Nuclear localization of DAF-16 returned to basal levels at 48 hours. This raises the possibility that DAF-16 activity during that window of time is sufficient for long-term protection by means of the stability of its transcriptional targets. Strikingly, although a daf-16 mutation did not cause TRN loss of integrity in wild-type animals, it completely abolished the protection by HT115 diet on mec-4d mutants. Taken together, these results show that DAF-16 is critical under conditions of neuronal stress and necessary for the dietary protection mediated by E . coli HT115 to take place.

One of the metabolites that are overabundant in HT115 compared with OP50 is GABA. Interestingly, loss of the GAD enzyme, which catalyzes GABA production, diminished DAF-16 translocation in HT115 bacteria. GABA alone, however, supplemented to E . coli OP50 did not promote nuclear localization of DAF-16. These results suggest that although GABA is required for DAF-16 localization, it is not sufficient by itself, and other metabolites produced by HT115, such as lactate, act cooperatively to produce this translocation. Alternatively, this could suggest that although DAF-16 expression is required for neuroprotection to occur, neuroprotective metabolites likely function through a mechanism that does not involve active translocation of DAF-16 above basal levels. Currently, we cannot discriminate between these two possibilities, and more research is needed to further clarify the molecular pathways involved in GABA- mediated effects in neuronal integrity preservation by HT115. This work reveals a complex scenario of communication between bacteria and host that is likely to involve systemic and neuron-specific changes.

Materials and methods C . elegans maintenance and growth

Wild-type (N2) and mutant strains TU2773 ( uIs31 [ P mec-17 mec-17 :: gfp ] mec-4d [ e1611 ]), CF1139 ( daf-16 [ mu86 ] muIs61 [ P daf-16 daf-16 :: gfp ] rol-6 [ su1006 ]), WCH34 ( daf-2ts [ e1368 ] mec-4d [ e1611 ] uIs31 [ P mec-17 mec-17 :: gfp ]), WCH39 ( daf-16 [ m27 ] mec-4d [ e1611 ] uIs31 [ P mec-17 mec-17 :: gfp ]), WCH40 ( daf-16 [ m27 ] uIs31 [ P mec-17 mec-17 :: gfp ]), TU38 ( deg-1 [ u38 ]), TU3755 ( uIs58 [ P mec-4 mec-4 :: gfp ]), WCH6 ( uIs71 [ P mec-18 sid-1Pmyo-2mcherry ], uIs31 [ Pmec-17mec-17 :: gfp ], sid-1 [ pk3321 ], mec-4d [ e1611 ]), and WCH41 ( uIs58 [ P mec-4 mec-4 :: gfp ] mec-4d [ e1611 ]) were grown at 20 °C as previously described [ 49 ]. All nematode strains are maintained on E . coli OP50 strain prior to feeding with other bacteria. Unless otherwise noted, all plates were incubated at 20 °C.

Bacteria were grown overnight on Luria-Bertani (LB) plates at 37 °C from glycerol stocks. The next morning, a large amount of the bacterial lawn was inoculated in LB broth and grown for 6 hours on agitation at 450 g at 37 °C. In all, 100 mL of this bacterial culture was seeded onto 60-mm NGM plates and allowed to dry overnight before worms were placed on them. We used the following bacterial strains as worm food: E . coli OP50, E . coli HT115, E . coli K-12 BW25113, E . coli B BL21 DE3, C . aquatica , C . testosteroni , B . megaterium , P . aeruginosa PAO1, Pseudochrobactrum sp., Stenotrophomonas sp., B . pumilus , and E . coli HT115 Δ gad .

Pertinence of the strains selected

We chose a diverse array of bacterial strains that have been used before in C . elegans research as well as bacteria from a wild microbiome. The rationale for using laboratory strains is (1) their availability and (2) further characterization of strains currently used and whose effect in neuroprotection and metabolite production may be helpful in result interpretation for the larger C . elegans research community. Specifically, we used strains for routine laboratory work that served as a control ( E . coli OP50) strains for dsRNA delivery in C . elegans RNAi experiments ( E . coli HT115) the two E . coli strains from which OP50 and HT115 were derived ( E . coli B and K-12, respectively) a mild pathogen used to study bacteria–host interactions ( P . aeruginosa PAO1) C . aquatica , a high-quality food for C . elegans and B . megaterium , which causes extreme roaming and is hard to eat. Wild bacteria were included because they were isolated from wild C . elegans and constitute a natural microbiome.

UV killing of bacteria Killing on plates

Bacterial cultures (300 μL) grown at an OD of 0.8 were inoculated on 60-mm NGM plates. Once the liquid dried (overnight), plates were placed upside down in the UV transilluminator (Cole-Palmer HP 312 nm) for 5 minutes at high power. Synchronized animals were seeded onto the plates immediately after bacterial killing. An inoculum of the UV-treated bacterial lawn was picked and streaked onto new LB plates to confirm that bacteria were effectively killed.

Bacterial cultures (5 mL) grown at an OD of 0.8 were placed in an empty 60-mm plate and exposed to UV light for 10 minutes with slow agitation every 5 minutes. We used 300 μL of this liquid to seed 60-mm plates.

Criteria for neuronal integrity AVM neuron

Morphological evaluation of AVM neuron was modified from Calixto and colleagues, 2012 [ 16 ]. Neurons with full-length axons, as well as those with anterior processes that passed the point of bifurcation to the nerve ring, were classified as AxW (see Fig 1A ). Axons with a process connected to the nerve ring were classified as AxL, and those that did not reach the bifurcation to the nerve ring were classified as AxT. Lack of axon, only soma, and soma with only the ventral projection were classified as Axϕ.

Neurons with full-length axons were classified as AxW. AxT were ALM neurons with axons that did not reach the bifurcation to the nerve ring, and PLMs were axons that did not reach mid body. Neurons without axons or somas were classified as Axϕ.

For morphological evaluation, worms were mounted on 2% agarose pads, paralyzed with 1 mM levamisole, and visualized under a Nikon Eclipse Ti-5 fluorescence microscope with 40× or 60× magnification under Nomarski optics or fluorescence. For high-resolution images, we used a Leica TCS SP5X microscope. DAF-16 nuclear expression and MEC-4 localization in CF1139 and TU3755 animals, respectively, were quantified using ImageJ (1.46v). For accuracy in the categorization and to avoid damage due to long exposure to levamisole, animals were scored within 20 minutes after placing them on the agarose pads.

Clones of interest were taken from the Ahringer RNAi bacterial library [ 50 , 51 ]. We performed a P0 screen as described in Calixto and colleagues [ 16 ] in mec-4d animals and TRN-autonomous RNAi strain carrying the mec-4d ( e1611 ) mutation (WCH6, [ 16 ]). Bacterial clones were taken from glycerol stocks and grown overnight on LB plates containing tetracycline (12.5 μg/mL) and ampicillin (50 μg/mL). The next morning, a chunk of bacterial lawn was grown on liquid LB containing ampicillin (50 μg/mL) for 8 hours. A total of 400 μL of bacterial growth was plated on NGM plates containing 1 mM IPTG and carbenicillin (25 μg/mL) and allowed to dry until the next day, when 30–50 newly hatched mec-4d or WCH6 worms were placed on NGM plates. ALM integrity was scored in young adults 72 hours later.

Synchronization of animals

Plates with large amounts of laid eggs were washed with M9 to eliminate all larvae and adults. Within the next 2 hours, newly hatched L1 animals were collected with a mouth pipette and transferred to the desired experimental plates.

Time course of neuronal degeneration

Synchronized L1 larvae were placed in plates at 20 °C with the desired bacterial food using a mouth pipette. We scored the integrity of the AVM neuron axon (1) during development at 12, 24, 48, and 72 hours posthatching (2) every 24 hours for longer periods (from 0 or 12 hours until 168 hours posthatching) and (3) at adulthood (72 hours) after each treatment. The same experiments at 25 °C also included ALM and PLM neurons. For experiments at 25 °C without temperature shifts, parents of animals examined were kept at the same temperature starting as L4s. Animals were scored until 168 hours at intervals of 24 hours. For each evaluation, we used at least three biological replicas with triplicates of 30 worms each.

Synchronized L1 animals were placed in NGM plates seeded with UV-killed E . coli HT115 or OP50 for 6 or 12 hours and later changed to E . coli OP50 or HT115 until 72 hours posthatching. To transfer 6- and 12-hour-old animals to the new bacteria, larvae were washed off from the plates with sterile water supplemented with carbenicillin (25 μg/mL). Each replica was collected using a pipette in an Eppendorf tube. Animals were subsequently centrifuged for 2 minutes at 450 g . The pellet was washed with sterile water supplemented with carbenicillin followed by centrifugation. After two washes, the pellet was collected and transferred to new plates. Axonal morphology was evaluated at 12, 24, 48, and 72 hours. For controls in both experiments, we grew synchronized animals in the same way on E . coli OP50 and HT115 for their entire development until 72 hours.

DAF-16 localization on different foods

The 30 L4 CF1139 ( daf-16 [ mu86 ] muIs61 [ P daf-16 daf-16 :: gfp ] rol-6 [ su1006 ]) animals were placed in plates seeded with E . coli OP50, wild-type HT115, HT115 Δ gad mutant, or E . coli supplemented with 2 mM lactate or 2 mM GABA and allowed to lay eggs for 24 hours. The 30 L1 larvae, synchronized 0–2 hours posthatching, were transferred with a mouth pipette to new plates with E . coli HT115 or OP50. Morphological evaluation was performed at 12, 24, and 48 hours posthatching.

We quantified GFP-positive nuclei in whole body. The most visible expression was in intestinal and hypodermal cells.

MEC-4 puncta quantification at different temperatures

The 30 L4 TU3755 ( uIs58 [ P mec-4 mec-4 :: gfp ]) worms were fed E . coli OP50 or HT115 and allowed to lay embryos for 24 hours. The 30–50 L1 larvae were synchronized as described above and placed on the corresponding diet at 15, 20, or 25 °C.

Criteria for DAF-16 nuclear localization

We counted the number of GFP-positive nuclei between the terminal pharyngeal bulb and 30 μm after the vulva (or developing vulva). Because we performed a time course evaluation, animals had different body sizes at each time point. The comparisons were made between animals of the same life stage feeding on the two bacteria. For 12-hour-old animals, pictures were taken in a 60× objective, and for 24- and 48-hour-old animals, pictures were taken in a 40× objective. Experiments were done in triplicate for each time point.

Intensity of nuclear expression was estimated as “high” if there were more than 20 GFP-positive nuclei, “medium” between 10 and 19, “low” between 3 and 9, and “undetectable” if there were fewer than 3.

MEC-4 puncta quantification

One PLM per L4 animal was photographed under a 40× objective. The number of puncta was counted in 100 μm of axon starting from the neuronal soma. Images were visualized in ImageJ (1.46v), and puncta were counted manually.

To evaluate the functionality of the AVM mechanoreceptor neuron and PVC interneuron, the ability of animals to respond to gentle touch was tested. Animals were touched at 20 °C.

First, animals were synchronized in L1 larvae and placed 30 per NGM plate seeded with different bacteria (mentioned above). Then, in the case of AVM neuron, animals were touched with an eyebrow one time in the head, gently stroking where the pharyngeal bulb lies at 72 hours, whereas for the PVC interneuron, animals were gently touched with an eyebrow 10 times in a head-to-tail fashion every 24 hours after hatching.

To evaluate the effects of DAF-2 down-regulation on dauer animals, we used the strains WCH34 ( daf-2ts [ e1368 ] uIs31 [ P mec-17 mec-17 :: gfp ] mec-4d [ e1611 ]) and TU2773 ( uIs31 [ P mec-17 mec-17 :: gfp ] mec-4d [ e1611 ]) as a control. L4 animals from both strains were placed on plates seeded with E . coli HT115 and OP50 at 25 and 20 °C. Then, animals were synchronized by taking 30–50 L1 larvae and placing them in three plates for each point of evaluation. Morphology of TRNs (AVM, PLMs, ALMs, and PVM) was scored from 12 to 168 hours with intervals of 24 hours.

Longitudinal analysis of neuronal degeneration

At hatching (0–2 hours), the 30 synchronized L1s were placed on individual plates seeded with E . coli HT115. Animals were examined every 24 hours for 3 days, starting at 24 hours posthatching, by placing them on 2 μl of 0.1-μm polystyrene beads for AVM observation and photography. We used polystyrene beads in order to maintain the shape of the animal, allowing its rescue from the agarose pad for posterior visualizations. Each animal was gently returned in M9 to the plate with a mouth pipette.

Bacterial growth with controlled OD

E . coli HT115 bacteria were inoculated in LB starting from a –80 °C glycerol stock and allowed to grow for 1 hour. Nine different falcon tubes were used to grow bacteria. After the first hour, 1 mL from each tube was taken to measure the OD using a spectrophotometer (Ultraspec 2100). When the cultures reached the desired OD, growth was stopped. This procedure was repeated for each measurement. To obtain the lowest values (0.4/0.6/0.8), cultures were evaluated every hour and, for the rest of the values (0.8–2.0), every 15 or 30 minutes. A total of 200 μL of each culture was inoculated onto six NGM plates. Seeded plates were dried on a laminar flow hood for at least 1 hour. Finally, bacteria were killed on a UV transilluminator (Cole-Palmer High performance) by placing the open plate upside down for 5 minutes using the highest wavelength (365 nm). On three of those plates, 30 synchronized L1 worms were placed. The other three plates were used to test whether bacteria were killed during the protocol. To this end, parts of the lawn of UV-killed bacteria were streaked onto another LB plate to observe whether bacteria grew on the agar within the next 3 days.

Supplementation with bacterial supernatant

Overnight E . coli OP50 and HT115 bacterial cultures (5 mL) were centrifuged for 15 minutes at 3,500 g . The supernatant was sieved twice on 0.2-μm filters using a sterile syringe to separate bacteria in suspension. E . coli OP50 supernatant was used to resuspend E . coli HT115 pellet, and the same was done with E . coli HT115 supernatant on OP50 pellet. Each bacterial suspension was inoculated on NGM plates as described above.

Supplementation with GABA, glutamic acid, or lactate

After liquid UV killing, the desired bacteria were mixed with 2 mM GABA, 2 mM glutamate, or 2 mM of L-lactate (Sigma). Bacterial cultures (400 μL) were inoculated in 60-mm NGM plates to cover the entire surface. The next day, 0- to 2-hour synchronized L1 worms were placed in the NGM plates for 24 hours. Animals were moved to freshly prepared plates every 24 hours for 72 hours.

Mix of bacteria and UV killing

Bacterial cultures were combined in different proportions with the other bacteria or LB for controls. Liquid LB media growth until OD 0.6 of E . coli OP50 or HT115 were diluted 1:125 in LB and incubated at 37 °C with shaking O/N. Bacterial cultures were mixed in the indicated proportions and used to prepare plates as described before (0.1%, 1%, 10%, and 50% of E . coli HT115 in OP50). Plates with mixed bacteria were irradiated with UV before adding the worms.

Generation of bacterial gad mutant

Two orthologs of the gad gene are present in E . coli HT115, denominated gadA and gadB . E . coli HT115 mutants were constructed using homologous recombination with PCR products as previously reported [ 52 ]. Wild-type E . coli HT115 [ 53 ] was transformed with plasmid pKD46 [ 52 ]. Next, electrocompetent bacteria were prepared at 30 °C in SOB medium with ampicillin (100 μg/mL) and arabinose and electroporated with a PCR product obtained using the set of primers gadBH1P1 and gadBH2P2 and pKD3 as template. Recombinant candidates were selected in LB plates plus chloramphenicol at 37 °C. Colonies were tested for loss of ampicillin resistance. Amp s colonies were checked for substitution of the gadB gene by the chloramphenicol acetyl transferase cassette by PCR using primers gadB-A and gadB-B, which flank the insertion site. This rendered the E . coli HT115 Δ gadB :: cat derivative.

This strain was electroporated with pKD46. Competent cells of this transformed strain were prepared with ampicillin and arabinose and electroporated with a PCR product generated using primers NgadAH1P1 and NgadAH2P2 and pKD4 as template.

Recombinant candidates were selected in LB plates plus kanamycin (30 μg/mL) at 37 °C. Colonies obtained were tested for loss of ampicillin resistance and checked for substitution of the gadA gene by the kanamycin resistance gene through PCR using primers NGadAFw and NgadARv flanking the substitution site. This yielded an E . coli HT115 Δ gadB :: cat /Δ gadA :: kan double mutant strain (referred as E . coli HT115Δ gad mutant in the text). All primers are listed in S3 Table .

Generation of the pG gadA plasmid

E . coli HT115 gadA gene was amplified by PCR using primers NgadAFw and NgadARw with Taq polymerase and cloned in the pGemT Easy plasmid (Promega) according to the manufacturer’s instruction to generate pG gadA . E . coli OP50 and E . coli HT115 Δ gadB :: cat /Δ gadA :: kan were transformed chemically with the pG gadA plasmid. Competent bacteria were obtained using a calcium chloride protocol [ 54 ]. Transformed colonies for E . coli OP50 + pG gadA and HT115 Δ gadB :: cat /Δ gadA ::kan+pG gadA were confirmed by antibiotic resistance (ampicillin 100 μg/mL), plasmid purification, plasmid length, and gene endonuclease restriction with XbaI (NEB).

Glutamate decarboxylase enzymatic activity was measured according to Rice and colleagues [ 29 ], with modifications [ 28 ]. Fresh bacterial colonies were grown in LB with the appropriate antibiotic until 108 CFU/mL (3–4 hours). Antibiotics used were 25 μg/mL of streptomycin for E . coli OP50, 25 μg/mL of tetracycline for HT115, 20 μg/mL of kanamycin and chloramphenicol for HT115Δ gad , and 25 μg/mL of streptomycin and ampicillin and 25 μg/mL of IPTG for OP50 + pG gadA .

Then, 10 mL of each culture was centrifuged at 500 g for 10 minutes, and the pellet was resuspended in 5 mL of phosphate buffer (KPO4 1M [pH 6.0]). This step was repeated one more time, and the pellet was resuspended in 2 mL of GAD reagent (1 g glutamic acid [Sigma], 3 mL Triton X-100 [Winkler], 0.05 g bromocresol blue [Winkler], 90 g sodium chloride [Winkler] for 1 L of distilled water with final pH 3.4) and preserved at 4 °C for 2 months maximum. Once the pellet was resuspended, the samples were measured for colorimetric differences (UltraSpect 2100) at 620 nm. Liquid LB treated without bacteria is used as a blank. Average colorimetric value for E . coli HT115 was considered 100% of possible GAD activity for each replicate.

Growth of bacterial cultures for GABA quantification by GABase assay

Bacterial cultures were grown in LB media with 25 μg/mL of streptomycin for E . coli OP50, 25 μg/mL of tetracycline for HT115, and 20 μg/mL of kanamycin and chloramphenicol for HT115Δ gad until 10 7 CFU/mL (3–4 hours).

Bacterial GABA quantification by GABase assay

This reaction consisted of two enzymes that convert a molecule of GABA to succinate by GABA transaminase (GAT) and SSA dehydrogenase (SADH), producing detectable NADPH at 340 nm. Additionally, by inhibition of GAT with aminoethyl hydrogen sulfate, substrate concentrations for each reaction are distinguishable according to OD [NADPH] total = OD [NADPH] GAT + OD [NADPH] SADH.

Bacterial GABA quantification

Measurement of GABA was performed following the GABase Sigma protocol with modifications [ 30 ].

Reaction was performed in 100-μl final volume containing 1U/mL of GABase previously mixed with 75 mM of potassium phosphate buffer in 25% of glycerol, 100 mM of pyrophosphate potassium buffer, 500 μM of NADP+, 5 mM of a-ketoglutarate, 100 μM of dithiothreitol, 50 mM of aminoethyl sulfate as GAT inhibitor, and 10 uL of each sample. All reactions were performed in 96-well Nunclon plates in a NanoQuant 200 Infinite spectrophotometer for 1 hour at 340 nm and 37 °C. Every run contained positive (5 mM of GABA) and negative controls (no GABAse or no GABA) to be subtracted as background of total NADPH measurement, as well as inhibited reaction with respective controls. The GABA concentration curve was performed in triplicate to measure absorbance from 0 to 5 mM of GABA (Abs = 0.1138 × [GABA] + 0.04522).

GABA extraction for GABase reaction

All bacteria culture extractions were performed from 1 × 10 8 CFU/mL three times in each triplicate. A total of 1 mL of culture was centrifuged at 1,100 g (4,000 rpm) for 10 minutes, and pellet was washed twice with phosphate buffer (1 M KPO4 [pH 6.0]) by pipetting and centrifugation at 3,300 g (7,000 rpm) for 10 minutes. After the last centrifugation, 100 μL of deionized water was added, and samples were treated in a water bath at 95 °C for 15 minutes. Then, a final centrifugation at 1,100 g (4,000 rpm) for 10 minutes separated GABA in the supernatant from bacteria debris. Extracted samples were frozen at –20 °C until immediate use in the reaction the next day.

Neuronal morphology for mec-4d worms, GAD activity, and GABA production in each bacterium were correlated using one-tailed Pearson’s correlation analysis, and linear regression confirmed the variables’ slope differences.

Growth of bacteria for NMR spectroscopy

Bacterial strains were grown in solid LB media with 25 μg/mL of streptomycin for E . coli OP50, 25 μg/mL of tetracycline for E . coli HT115, and 20 μg/mL of kanamycin and chloramphenicol for HT115Δ gad at 37 °C for 10 hours. Eight preinocula of 2 mL of liquid LB were set for each bacterial strain, grown overnight, and continued in 35 mL of LB and monitored by OD until required (OD of 1).

Metabolite extraction for NMR

Bacterial cultures were centrifuged at 4,000 g for 5 minutes to separate bacteria from the media. Bacterial pellet was resuspended in phosphate-buffered saline (PBS: NaCl [137 mM], KCl [2.7 mM], Na2HPO4 [10 mM], and KH2PO4 [1.8 mM]) and washed twice after centrifugation with PBS at 4,000 g for 5 minutes. After the last wash, each pellet was resuspended in 1 mL of cold extraction buffer and pipetted into Eppendorf tubes. The extract solution was prepared by mixing equal volumes of acetonitrile and KH2PO4/NaH2PO4 (100 mM, pH 7.4).

Bacterial membrane permeabilization was performed in two steps. First, Eppendorf tubes were submerged in a liquid nitrogen bath for 2 minutes, defrosted at 4 °C, and vortexed for 30 seconds. This procedure was repeated three times. Secondly, tubes from the first step were sonicated in an ultrasonic water bath (Bioruptor UCD-200, Diagenode) for 15 cycles of 30 seconds on and 30 seconds off at full power. Finally, to obtain the metabolites, samples were centrifuged for 10 minutes at 8,000 g , and the supernatant was recovered. This step was repeated to optimize metabolite recovery. Samples were dried in a vacuum dryer (Savant) for 60 minutes at 50 °C and 300 g .

The 1H NMR spectroscopy and multivariate data analysis were performed at Plataforma Argentina de Biología Estructural y Metabolómica (PLABEM).

Sample preparation for 1H NMR spectroscopy

Samples were randomized and reconstituted in 600 μL of 100 mM Na + /K + buffer (pH 7.4) containing 0.005% TSP (sodium 3-trimethylsilyl- (2,2,3,3-2H4)-1-propionate) and 10% D2O. In order to remove any precipitate, samples were centrifugated for 10 minutes at 14,300 g at 4 °C. The 500 μL of the centrifuged solution was transferred into a 5-mm NMR tube (Wilmad LabGlass).

1H NMR spectroscopic analysis of bacterial extracts

NMR spectra were obtained at 300 K using a Bruker Avance III 700-MHz NMR spectrometer (Bruker Biospin, Rheinstetten, Germany) equipped with a 5-mm TXI probe. One-dimensional 1H NMR spectra of bacterial extracts were acquired using a standard 1-D noesy pulse sequence (noesygppr1d) with water presaturation [ 55 , 56 ]. The mixing time was set to 10 milliseconds, the data acquisition period to 2.228 seconds, and the relaxation delay to 4 seconds. The 1 H NMR spectra were acquired using four dummy scans and 32 scans with 64 K time domain points and a spectral window of 20 ppm. FIDs were multiplied by an exponential weighting function corresponding to a line broadening of 0.3 Hz. Two-dimensional NMR spectra 1H-1H TOCSY and 1H J-resolved pulse sequences were acquired for resonance assignment purposes.

Quality controls (QC) were prepared as suggested by Dona and colleagues [ 55 ]. The 1H NMR spectra of QC samples were acquired every eight study samples.

1H NMR spectral processing

Spectroscopic data were processed in Matlab (version R2015b, The MathWorks). Spectra were referenced to TSP at 0.0 ppm, and baseline correction and phasing of the spectra were achieved using Imperial College written functions (provided by T. Ebbels and H. Keun, Imperial College London). Each spectrum was reduced to a series of integrated regions of equal width (0.04 ppm, standard bucket width). Noninformative spectral regions containing no metabolite signals, TSP signal, and the interval containing the water signal (between 4.9 and 4.6 ppm) were excluded. Each spectrum was then normalized by the probabilistic quotient method [ 57 ]. Spectra alignment was made using the alignment algorithm recursive segment-wise peak alignment [ 58 ] in user-defined windows.

Statistical analysis of NMR spectroscopic data

The preprocessed 1H NMR spectral data were imported to SIMCA (version 14.1, Umetrics AB, Umeå, Sweden) for multivariate data analysis. PCA was performed on the Pareto-scaled NMR dataset. OPLS-DA was made to maximize the separation between bacterial groups as a function of neuroprotection. To ensure valid and reliable OPLS-DA models and to avoid overfitting, 200 permutations were carried out. Discriminant features between classes in OPLS-DA models were defined using a combination of loading plot (S-line plot) and VIP plot. Variables met highest p(corr)1 in S-Line plot, and VIP values >1 were selected and validated by spectral raw data examination.

NMR resonances assignment

Two-dimensional NMR spectra 1H J-resolved pulse sequences were acquired for resonance assignment purposes. Discriminant features were assigned searching in the E . coli Metabolome Database (ECMDB) [ 59 , 60 ]. The unequivocal identification of GABA and glutamate were made through spike-in experiments over an E . coli HT115 and HT115 Δ gad extract, respectively.

Growth of bacteria expressing gad plasmid

Genetically complemented bacteria with pG gadA were grown until an OD of 0.6. GAD expression was induced by adding 0.15 mM IPTG to OD-0.6 liquid cultures. After, 1-hour samples of each condition were taken to assess GAD activity and to seed on NGM plates with the appropriate antibiotic and supplemented with 0.1 mM of IPTG.

Genome sequencing of E . coli OP50 and HT115

E . coli OP50 and HT115 were grown from glycerol stock on LB plates overnight. The next morning, portions of the lawn were cultured on agitation for 4 hours in liquid LB. Liquid cultures (2 mL) were pelleted, and DNA was purified using the UltraClean microbial DNA isolation kit (MO BIO) according to the manufacturer’s instructions.

Whole-genome sequencing was done at Genome Mayor Sequencing Services. Paired-end reads (2 × 250 bp) were generated using the Illumina MiSeq platform. Sequence data were trimmed using Trimmomatic version 0.27 [ 61 ]. Trimmed reads were assembled using SPAdes version 3.1.0 [ 62 ]. Genome annotation for both organisms was done using PROKKA version 1.9 [ 63 ]. Both genomes were deposited to NCBI under the accession numbers PRJNA526029 ( E . coli OP50) and PRJNA526261 ( E . coli HT115).

Comparative genome analysis of E . coli OP50 and HT115 Identification of common and unique sequences of E. coli strains

Annotated proteins in fasta format from assembled genomes of E . coli OP50 (PRJNA526029) and HT115 (PRJNA526261) were compared between strains by Reciprocal Best Hit analysis using Blast+ [ 64 , 65 ]. We ran Blastp [ 66 ] of the protein sequences of OP50 strain against HT115. Then, we ran Blastp of sequences from the E . coli HT115 strain against OP50. We extracted top hits based on bit scores and E-values. Then, we compared both top hits and selected the best match of each other, named “common sequences.” All the other sequences were defined as “unique.”

For total RNA isolation, E . coli OP50 and HT115 were grown from glycerol stock on LB plates overnight. The next morning, portions of the lawn were cultured on agitation in liquid LB for 5 hours. Then, bacterial cultures (2 mL) were pelleted for RNA extraction with Max Bacterial Trizol kit (Invitrogen) according to the manufacturer’s protocol.

cDNA libraries for Illumina sequencing were generated by Centro de Genómica y Bioinformática, Universidad Mayor, Chile. cDNA libraries were made with Illumina Truseq stranded mRNA kit according to the manufacturer’s protocol in Illumina HiSeq platform. QC of libraries was made with bioanalyzer, and quantification was done with qPCR StepOnePlus Applied Biosystem. Six sets of lllumina paired-end reads in FASTQ format corresponding to three replicates from E . coli strain HT115 and three replicates from strain OP50 were analyzed as follows.

Data preprocessing and QC

Reads with an average quality lower than 30 over four bases, as well as reads shorter than 16 bp, were discarded with Trimmomatic version 0.35 [ 61 ]. Pre- and posttrimming quality visualization was made with FastQC ( https://www.bioinformatics.babraham.ac.uk/projects/fastqc/ ).

Mapping and quantification of transcript abundance

Mapping and quantification from E . coli HT115 and OP50 strains were made using Bowtie2 [ 67 ] with default parameters. We used as reference the ASM435494v1 assembly from E . coli HT115 (accession GCA_004354945.1) and ASM435501v1 from E . coli OP50 (GCA_004355015.1).

Differential expression analysis

This analysis was performed preserving only common sequences transcript abundance was compared between strains. For this, read quantification was performed with FeatureCounts in Rsubread [ 68 ] in R (version 3.5), and then differential expression analysis was performed using DeSeq2 version 3.8 [ 69 ] using default parameters. Cutoff for differentially expressed genes (DEG) was set at adjusted p -value (padj) < 0.01 and is reported in S2 File .

Criteria for gene expression level

We categorized gene expression according to [ 70 ] as follows: low if expression is between 0.5 and 10 FPKM or 0.5 and 10 TPM medium if expression is between 11 and 1,000 FPKM or 11 and 1,000 TPM and high if expression was more than 1,000 FPKM or 1,000 TPM.

Biological and technical replicates

Each experiment was performed in three technical triplicates and at least three biological replicates. We define biological replicates as experiments made on different days, containing triplicates of each condition, and a technical replicate as a triplicate of the same condition on the same day. The average of the three reads of each triplicate is considered as one count. Each experiment has three technical replicates that were in turn averaged to constitute one of the points of each figure. Data are collected and processed as a single technical replicate (the average of three counts of the same plate), and its mean is used as a single biological replicate. Each figure contains at least three experiments (biological replicates) performed as explained before. All the biological replicates are performed spaced from each other from 1 day to 1 week.

Each experiment started with at least 30 synchronized worms on each technical triplicate with the exception of the longitudinal study of degeneration on E . coli HT115, which used 30 animals in total.

Statistical evaluation was performed using one- or two-way ANOVA with post hoc tests and a Student t test when indicated. Results of all tests are detailed in S2 Dataset .

Supporting information S1 Fig Axonal categories at different optical density.

(A and B) All axonal categories (A) and wild-type axons (B) in worms feeding on E . coli HT115 bacteria grown to different optical density. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively.

S2 Fig Axonal morphological categories of different classes of neurons in E . coli HT115 and OP50.

All axonal categories of animals feeding on E . coli OP50 (A, C, and E) and HT115 (B, D, and F) at 25 °C. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively.

S3 Fig Quantification of mechanosensory channel MEC-4 expression in vivo .

Number of puncta in 100 μm of PLM axons on each bacterial diet (A) and representative photograph of PLM axons used for quantification. Size bar is 20 μm. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. MEC-4, mechanosensory ion channel subunit PLM, posterior lateral microtubule.

S4 Fig Complete axonal categories of priming experiments.

(A–D) All axonal categories of animals feeding E . coli HT115 for 6 (A) and 12 (B) hours with controls of ad libitum E . coli OP50 (C) and HT115 (D) or feeding E . coli OP50 for 6 (E) and 12 (F) hours with controls of ad libitum E . coli OP50 (G) and HT115 (H). The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively.

S5 Fig Complete axonal categories for animals feeding on modified bacteria.

(A and B) All axonal categories of animals feeding wild-type and Δ gad E . coli HT115 (A) and OP50 (B) modified with Gad-expressing plasmids, glutamate, and GABA. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. GABA, γ-aminobutyric acid gad , glutamate decarboxylase enzyme gene.

S6 Fig Multivariate analysis and model’s validation.

(A) PC score plot derived from 1H NMR spectra indicating metabolic differences between wild-type E . coli strains OP50 (orange) and HT115 (green) and HT115 Δ gad mutant (light gray). Quality controls are displayed in yellow. Model parameters are R2X = 0.787 and Q2 = 0.705. (B and C) OPLS-DA validation by 200 permutations. E . coli HT115 and HT115 Δ gad validate model intercepts: R2 = (0.0 0.656) and Q2 = (0.0 −0.626) (B). Protective E . coli HT115 and nonprotective strains E . coli OP50 and HT115 Δ gad validate model intercepts: R2 = (0.0 0.329) and Q2 = (0.0 −0.566) (C). The underlying numerical data for each figure panel can be found in S1 Dataset . gad , glutamate decarboxylase enzyme gene NMR, nuclear magnetic resonance OPLS-DA, orthogonal projections to latent structures discriminant analysis PC, principal component.

S7 Fig Resonances assignment confirmation by NMR.

(A and B) 1H NMR spectra of GABA (A) and glutamate (C). (B–D) Spike-in of GABA and glutamate confirms identity of metabolites. E . coli HT115 extract (red) (B) E . coli HT115 Δ gad extract (blue). Spike-in was made adding 5 μL of standard 10 mM twice. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Tables. GABA, γ-aminobutyric acid gad , glutamate decarboxylase enzyme gene NMR, nuclear magnetic resonance.

S8 Fig Protection conferred by lactate supplementation of E . coli OP50.

All axonal categories of mec-4d animals feeding on E . coli OP50 supplemented with lactate. The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. mec-4d , mechanosensory abnormality protein 4.

S9 Fig Standardization curve for GABA concentration.

Known GABA concentration plotted against absorbance values creates a curve for later estimation of GABA in samples. The underlying numerical data can be found in S1 Dataset . GABA, γ-aminobutyric acid.

S10 Fig All morphological categories of neuroprotection in animals treated with dsRNA for GABA effectors.

Complete morphological categories of mec-4d animals feeding on E . coli HT115 expressing dsRNA for GABA effector systemically (A) and touch neuron autonomously (B). The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. dsRNA, double-stranded RNA GABA, γ-aminobutyric acid mec-4d , mechanosensory abnormality protein 4.

S11 Fig Effect of DAF-2 down-regulation on neuronal degeneration in animals fed E . coli OP50.

(A–C) Neuronal integrity of AVM (A), ALM (B), and PLM (C) neurons of daf-2(ts ) mec-4d animals fed OP50. (D–E) All axonal categories of daf-16 mec-4d animals fed E . coli OP50 (D) and HT115 (E). The underlying numerical data and statistical analysis for each figure panel can be found in S1 and S2 Datasets, respectively. ALM, anterior lateral microtubule AVM, anterior ventral microtubule daf-2 , codes for insulin-like growth factor 1 (IGF-1) receptor daf-16 , ortholog of the Forkhead box transcription factor mec-4d , mechanosensory abnormality protein 4 PLM, posterior lateral microtubule.

S1 Table 1H NMR resonance assignments.

E . coli HT115 and HT115 Δ gad mutant extracts. D, doublet dd, double doublet gad , glutamate decarboxylase enzyme gene m, multiplet NMR, nuclear magnetic resonance S, singlet t, triplet.

S2 Table 1H NMR resonance assignments.

Neuroprotective (HT115) and no neuroprotective (OP50 and HT115) bacterial extracts. D, doublet dd, double doublet m, multiplet NMR, nuclear magnetic resonance S, singlet t, triplet.

S3 Table Primers used for the construction of the E . coli HT115 Δ gadB :: cat /Δ gadA :: kan double mutant strain.

gad , glutamate decarboxylase enzyme gene. cat , chloramphenicol resistance gene kan , kanamycin resistance gene.

S1 File Genomics data showing unique genes of E . coli OP50 and HT115.

S2 File Differential expression analysis for all shared genes between E . coli OP50 and HT115.

Contains sources of experimental materials, strains, and procedures used in this research. MDAR, Materials Design Analysis Reporting.

S1 Dataset All data with replicas from the experiments shown in the manuscript.

Numerical data for every experiment on the manuscript are contained in this dataset. Data for each figure are contained in independent tabs.

S2 Dataset All statistical analysis of the experiments contained in the manuscript.

Statistical and post hoc analysis for every experiment on the manuscript is contained in this dataset. Data for each figure are contained in independent tabs.


Materials and methods

C. elegans maintenance and growth

Bacterial growth

Bacteria were grown overnight on Luria-Bertani (LB) plates at 37 °C from glycerol stocks. The next morning, a large amount of the bacterial lawn was inoculated in LB broth and grown for 6 hours on agitation at 450g at 37 °C. In all, 100 mL of this bacterial culture was seeded onto 60-mm NGM plates and allowed to dry overnight before worms were placed on them. We used the following bacterial strains as worm food: E. coli OP50, E. coli HT115, E. coli K-12 BW25113, E. coli B BL21 DE3, C. aquatica, C. testosteroni, B. megaterium, P. aeruginosa PAO1, Pseudochrobactrum sp., Stenotrophomonas sp., B. pumilus, and E. coli HT115 Δgad.

Pertinence of the strains selected.

We chose a diverse array of bacterial strains that have been used before in C. elegans research as well as bacteria from a wild microbiome. The rationale for using laboratory strains is (1) their availability and (2) further characterization of strains currently used and whose effect in neuroprotection and metabolite production may be helpful in result interpretation for the larger C. elegans research community. Specifically, we used strains for routine laboratory work that served as a control (E. coli OP50) strains for dsRNA delivery in C. elegans RNAi experiments (E. coli HT115) the two E. coli strains from which OP50 and HT115 were derived (E. coli B and K-12, respectively) a mild pathogen used to study bacteria–host interactions (P. aeruginosa PAO1) C. aquatica, a high-quality food for C. elegans and B. megaterium, which causes extreme roaming and is hard to eat. Wild bacteria were included because they were isolated from wild C. elegans and constitute a natural microbiome.

UV killing of bacteria

Killing on plates.

Bacterial cultures (300 μL) grown at an OD of 0.8 were inoculated on 60-mm NGM plates. Once the liquid dried (overnight), plates were placed upside down in the UV transilluminator (Cole-Palmer HP 312 nm) for 5 minutes at high power. Synchronized animals were seeded onto the plates immediately after bacterial killing. An inoculum of the UV-treated bacterial lawn was picked and streaked onto new LB plates to confirm that bacteria were effectively killed.

Killing in liquid.

Bacterial cultures (5 mL) grown at an OD of 0.8 were placed in an empty 60-mm plate and exposed to UV light for 10 minutes with slow agitation every 5 minutes. We used 300 μL of this liquid to seed 60-mm plates.

Criteria for neuronal integrity

AVM neuron.

Morphological evaluation of AVM neuron was modified from Calixto and colleagues, 2012 [16]. Neurons with full-length axons, as well as those with anterior processes that passed the point of bifurcation to the nerve ring, were classified as AxW (see Fig 1A). Axons with a process connected to the nerve ring were classified as AxL, and those that did not reach the bifurcation to the nerve ring were classified as AxT. Lack of axon, only soma, and soma with only the ventral projection were classified as Axϕ.

ALM and PLM neurons.

Neurons with full-length axons were classified as AxW. AxT were ALM neurons with axons that did not reach the bifurcation to the nerve ring, and PLMs were axons that did not reach mid body. Neurons without axons or somas were classified as Axϕ.

Microscopy

For morphological evaluation, worms were mounted on 2% agarose pads, paralyzed with 1 mM levamisole, and visualized under a Nikon Eclipse Ti-5 fluorescence microscope with 40× or 60× magnification under Nomarski optics or fluorescence. For high-resolution images, we used a Leica TCS SP5X microscope. DAF-16 nuclear expression and MEC-4 localization in CF1139 and TU3755 animals, respectively, were quantified using ImageJ (1.46v). For accuracy in the categorization and to avoid damage due to long exposure to levamisole, animals were scored within 20 minutes after placing them on the agarose pads.

Feeding RNAi

Clones of interest were taken from the Ahringer RNAi bacterial library [50,51]. We performed a P0 screen as described in Calixto and colleagues [16] in mec-4d animals and TRN-autonomous RNAi strain carrying the mec-4d(e1611) mutation (WCH6, [16]). Bacterial clones were taken from glycerol stocks and grown overnight on LB plates containing tetracycline (12.5 μg/mL) and ampicillin (50 μg/mL). The next morning, a chunk of bacterial lawn was grown on liquid LB containing ampicillin (50 μg/mL) for 8 hours. A total of 400 μL of bacterial growth was plated on NGM plates containing 1 mM IPTG and carbenicillin (25 μg/mL) and allowed to dry until the next day, when 30–50 newly hatched mec-4d or WCH6 worms were placed on NGM plates. ALM integrity was scored in young adults 72 hours later.

Synchronization of animals

Plates with large amounts of laid eggs were washed with M9 to eliminate all larvae and adults. Within the next 2 hours, newly hatched L1 animals were collected with a mouth pipette and transferred to the desired experimental plates.

Time course of neuronal degeneration

Synchronized L1 larvae were placed in plates at 20 °C with the desired bacterial food using a mouth pipette. We scored the integrity of the AVM neuron axon (1) during development at 12, 24, 48, and 72 hours posthatching (2) every 24 hours for longer periods (from 0 or 12 hours until 168 hours posthatching) and (3) at adulthood (72 hours) after each treatment. The same experiments at 25 °C also included ALM and PLM neurons. For experiments at 25 °C without temperature shifts, parents of animals examined were kept at the same temperature starting as L4s. Animals were scored until 168 hours at intervals of 24 hours. For each evaluation, we used at least three biological replicas with triplicates of 30 worms each.

Food changes

Synchronized L1 animals were placed in NGM plates seeded with UV-killed E. coli HT115 or OP50 for 6 or 12 hours and later changed to E. coli OP50 or HT115 until 72 hours posthatching. To transfer 6- and 12-hour-old animals to the new bacteria, larvae were washed off from the plates with sterile water supplemented with carbenicillin (25 μg/mL). Each replica was collected using a pipette in an Eppendorf tube. Animals were subsequently centrifuged for 2 minutes at 450g. The pellet was washed with sterile water supplemented with carbenicillin followed by centrifugation. After two washes, the pellet was collected and transferred to new plates. Axonal morphology was evaluated at 12, 24, 48, and 72 hours. For controls in both experiments, we grew synchronized animals in the same way on E. coli OP50 and HT115 for their entire development until 72 hours.

DAF-16 localization on different foods

The 30 L4 CF1139 (daf-16[mu86] muIs61[Pdaf-16daf-16::gfp] rol-6[su1006]) animals were placed in plates seeded with E. coli OP50, wild-type HT115, HT115 Δgad mutant, or E. coli supplemented with 2 mM lactate or 2 mM GABA and allowed to lay eggs for 24 hours. The 30 L1 larvae, synchronized 0–2 hours posthatching, were transferred with a mouth pipette to new plates with E. coli HT115 or OP50. Morphological evaluation was performed at 12, 24, and 48 hours posthatching.

We quantified GFP-positive nuclei in whole body. The most visible expression was in intestinal and hypodermal cells.

MEC-4 puncta quantification at different temperatures

The 30 L4 TU3755 (uIs58 [Pmec-4mec-4::gfp]) worms were fed E. coli OP50 or HT115 and allowed to lay embryos for 24 hours. The 30–50 L1 larvae were synchronized as described above and placed on the corresponding diet at 15, 20, or 25 °C.

Criteria for DAF-16 nuclear localization

We counted the number of GFP-positive nuclei between the terminal pharyngeal bulb and 30 μm after the vulva (or developing vulva). Because we performed a time course evaluation, animals had different body sizes at each time point. The comparisons were made between animals of the same life stage feeding on the two bacteria. For 12-hour-old animals, pictures were taken in a 60× objective, and for 24- and 48-hour-old animals, pictures were taken in a 40× objective. Experiments were done in triplicate for each time point.

Intensity of nuclear expression was estimated as “high” if there were more than 20 GFP-positive nuclei, “medium” between 10 and 19, “low” between 3 and 9, and “undetectable” if there were fewer than 3.

MEC-4 puncta quantification

One PLM per L4 animal was photographed under a 40× objective. The number of puncta was counted in 100 μm of axon starting from the neuronal soma. Images were visualized in ImageJ (1.46v), and puncta were counted manually.

Touch response

To evaluate the functionality of the AVM mechanoreceptor neuron and PVC interneuron, the ability of animals to respond to gentle touch was tested. Animals were touched at 20 °C.

First, animals were synchronized in L1 larvae and placed 30 per NGM plate seeded with different bacteria (mentioned above). Then, in the case of AVM neuron, animals were touched with an eyebrow one time in the head, gently stroking where the pharyngeal bulb lies at 72 hours, whereas for the PVC interneuron, animals were gently touched with an eyebrow 10 times in a head-to-tail fashion every 24 hours after hatching.

Temperature shifts

To evaluate the effects of DAF-2 down-regulation on dauer animals, we used the strains WCH34 (daf-2ts[e1368] uIs31 [Pmec-17mec-17::gfp] mec-4d[e1611]) and TU2773 (uIs31[Pmec-17mec-17::gfp]mec-4d[e1611]) as a control. L4 animals from both strains were placed on plates seeded with E. coli HT115 and OP50 at 25 and 20 °C. Then, animals were synchronized by taking 30–50 L1 larvae and placing them in three plates for each point of evaluation. Morphology of TRNs (AVM, PLMs, ALMs, and PVM) was scored from 12 to 168 hours with intervals of 24 hours.

Longitudinal analysis of neuronal degeneration

At hatching (0–2 hours), the 30 synchronized L1s were placed on individual plates seeded with E. coli HT115. Animals were examined every 24 hours for 3 days, starting at 24 hours posthatching, by placing them on 2 μl of 0.1-μm polystyrene beads for AVM observation and photography. We used polystyrene beads in order to maintain the shape of the animal, allowing its rescue from the agarose pad for posterior visualizations. Each animal was gently returned in M9 to the plate with a mouth pipette.

Bacterial growth with controlled OD

E. coli HT115 bacteria were inoculated in LB starting from a –80 °C glycerol stock and allowed to grow for 1 hour. Nine different falcon tubes were used to grow bacteria. After the first hour, 1 mL from each tube was taken to measure the OD using a spectrophotometer (Ultraspec 2100). When the cultures reached the desired OD, growth was stopped. This procedure was repeated for each measurement. To obtain the lowest values (0.4/0.6/0.8), cultures were evaluated every hour and, for the rest of the values (0.8–2.0), every 15 or 30 minutes. A total of 200 μL of each culture was inoculated onto six NGM plates. Seeded plates were dried on a laminar flow hood for at least 1 hour. Finally, bacteria were killed on a UV transilluminator (Cole-Palmer High performance) by placing the open plate upside down for 5 minutes using the highest wavelength (365 nm). On three of those plates, 30 synchronized L1 worms were placed. The other three plates were used to test whether bacteria were killed during the protocol. To this end, parts of the lawn of UV-killed bacteria were streaked onto another LB plate to observe whether bacteria grew on the agar within the next 3 days.

Supplementation with bacterial supernatant

Overnight E. coli OP50 and HT115 bacterial cultures (5 mL) were centrifuged for 15 minutes at 3,500g. The supernatant was sieved twice on 0.2-μm filters using a sterile syringe to separate bacteria in suspension. E. coli OP50 supernatant was used to resuspend E. coli HT115 pellet, and the same was done with E. coli HT115 supernatant on OP50 pellet. Each bacterial suspension was inoculated on NGM plates as described above.

Supplementation with GABA, glutamic acid, or lactate

After liquid UV killing, the desired bacteria were mixed with 2 mM GABA, 2 mM glutamate, or 2 mM of L-lactate (Sigma). Bacterial cultures (400 μL) were inoculated in 60-mm NGM plates to cover the entire surface. The next day, 0- to 2-hour synchronized L1 worms were placed in the NGM plates for 24 hours. Animals were moved to freshly prepared plates every 24 hours for 72 hours.

Mix of bacteria and UV killing

Bacterial cultures were combined in different proportions with the other bacteria or LB for controls. Liquid LB media growth until OD 0.6 of E. coli OP50 or HT115 were diluted 1:125 in LB and incubated at 37 °C with shaking O/N. Bacterial cultures were mixed in the indicated proportions and used to prepare plates as described before (0.1%, 1%, 10%, and 50% of E. coli HT115 in OP50). Plates with mixed bacteria were irradiated with UV before adding the worms.

Generation of bacterial gad mutant

Two orthologs of the gad gene are present in E. coli HT115, denominated gadA and gadB. E. coli HT115 mutants were constructed using homologous recombination with PCR products as previously reported [52]. Wild-type E. coli HT115 [53] was transformed with plasmid pKD46 [52]. Next, electrocompetent bacteria were prepared at 30 °C in SOB medium with ampicillin (100 μg/mL) and arabinose and electroporated with a PCR product obtained using the set of primers gadBH1P1 and gadBH2P2 and pKD3 as template. Recombinant candidates were selected in LB plates plus chloramphenicol at 37 °C. Colonies were tested for loss of ampicillin resistance. Amp s colonies were checked for substitution of the gadB gene by the chloramphenicol acetyl transferase cassette by PCR using primers gadB-A and gadB-B, which flank the insertion site. This rendered the E. coli HT115 ΔgadB::cat derivative.

This strain was electroporated with pKD46. Competent cells of this transformed strain were prepared with ampicillin and arabinose and electroporated with a PCR product generated using primers NgadAH1P1 and NgadAH2P2 and pKD4 as template.

Recombinant candidates were selected in LB plates plus kanamycin (30 μg/mL) at 37 °C. Colonies obtained were tested for loss of ampicillin resistance and checked for substitution of the gadA gene by the kanamycin resistance gene through PCR using primers NGadAFw and NgadARv flanking the substitution site. This yielded an E. coli HT115 ΔgadB::catgadA::kan double mutant strain (referred as E. coli HT115Δgad mutant in the text). All primers are listed in S3 Table.

Generation of the pGgadA plasmid

E. coli HT115 gadA gene was amplified by PCR using primers NgadAFw and NgadARw with Taq polymerase and cloned in the pGemT Easy plasmid (Promega) according to the manufacturer’s instruction to generate pGgadA. E. coli OP50 and E. coli HT115 ΔgadB::catgadA::kan were transformed chemically with the pGgadA plasmid. Competent bacteria were obtained using a calcium chloride protocol [54]. Transformed colonies for E. coli OP50 + pGgadA and HT115 ΔgadB::catgadA::kan+pGgadA were confirmed by antibiotic resistance (ampicillin 100 μg/mL), plasmid purification, plasmid length, and gene endonuclease restriction with XbaI (NEB).

GAD enzymatic activity

Glutamate decarboxylase enzymatic activity was measured according to Rice and colleagues [29], with modifications [28]. Fresh bacterial colonies were grown in LB with the appropriate antibiotic until 108 CFU/mL (3–4 hours). Antibiotics used were 25 μg/mL of streptomycin for E. coli OP50, 25 μg/mL of tetracycline for HT115, 20 μg/mL of kanamycin and chloramphenicol for HT115Δgad, and 25 μg/mL of streptomycin and ampicillin and 25 μg/mL of IPTG for OP50 + pGgadA.

Then, 10 mL of each culture was centrifuged at 500g for 10 minutes, and the pellet was resuspended in 5 mL of phosphate buffer (KPO4 1M [pH 6.0]). This step was repeated one more time, and the pellet was resuspended in 2 mL of GAD reagent (1 g glutamic acid [Sigma], 3 mL Triton X-100 [Winkler], 0.05 g bromocresol blue [Winkler], 90 g sodium chloride [Winkler] for 1 L of distilled water with final pH 3.4) and preserved at 4 °C for 2 months maximum. Once the pellet was resuspended, the samples were measured for colorimetric differences (UltraSpect 2100) at 620 nm. Liquid LB treated without bacteria is used as a blank. Average colorimetric value for E. coli HT115 was considered 100% of possible GAD activity for each replicate.

Growth of bacterial cultures for GABA quantification by GABase assay

Bacterial cultures were grown in LB media with 25 μg/mL of streptomycin for E. coli OP50, 25 μg/mL of tetracycline for HT115, and 20 μg/mL of kanamycin and chloramphenicol for HT115Δgad until 10 7 CFU/mL (3–4 hours).

Bacterial GABA quantification by GABase assay

This reaction consisted of two enzymes that convert a molecule of GABA to succinate by GABA transaminase (GAT) and SSA dehydrogenase (SADH), producing detectable NADPH at 340 nm. Additionally, by inhibition of GAT with aminoethyl hydrogen sulfate, substrate concentrations for each reaction are distinguishable according to OD [NADPH] total = OD [NADPH] GAT + OD [NADPH] SADH.

Bacterial GABA quantification.

Measurement of GABA was performed following the GABase Sigma protocol with modifications [30].

Reaction was performed in 100-μl final volume containing 1U/mL of GABase previously mixed with 75 mM of potassium phosphate buffer in 25% of glycerol, 100 mM of pyrophosphate potassium buffer, 500 μM of NADP+, 5 mM of a-ketoglutarate, 100 μM of dithiothreitol, 50 mM of aminoethyl sulfate as GAT inhibitor, and 10 uL of each sample. All reactions were performed in 96-well Nunclon plates in a NanoQuant 200 Infinite spectrophotometer for 1 hour at 340 nm and 37 °C. Every run contained positive (5 mM of GABA) and negative controls (no GABAse or no GABA) to be subtracted as background of total NADPH measurement, as well as inhibited reaction with respective controls. The GABA concentration curve was performed in triplicate to measure absorbance from 0 to 5 mM of GABA (Abs = 0.1138 × [GABA] + 0.04522).

GABA extraction for GABase reaction.

All bacteria culture extractions were performed from 1 × 10 8 CFU/mL three times in each triplicate. A total of 1 mL of culture was centrifuged at 1,100g (4,000 rpm) for 10 minutes, and pellet was washed twice with phosphate buffer (1 M KPO4 [pH 6.0]) by pipetting and centrifugation at 3,300g (7,000 rpm) for 10 minutes. After the last centrifugation, 100 μL of deionized water was added, and samples were treated in a water bath at 95 °C for 15 minutes. Then, a final centrifugation at 1,100g (4,000 rpm) for 10 minutes separated GABA in the supernatant from bacteria debris. Extracted samples were frozen at –20 °C until immediate use in the reaction the next day.

Correlation analysis.

Neuronal morphology for mec-4d worms, GAD activity, and GABA production in each bacterium were correlated using one-tailed Pearson’s correlation analysis, and linear regression confirmed the variables’ slope differences.

Growth of bacteria for NMR spectroscopy

Bacterial strains were grown in solid LB media with 25 μg/mL of streptomycin for E. coli OP50, 25 μg/mL of tetracycline for E. coli HT115, and 20 μg/mL of kanamycin and chloramphenicol for HT115Δgad at 37 °C for 10 hours. Eight preinocula of 2 mL of liquid LB were set for each bacterial strain, grown overnight, and continued in 35 mL of LB and monitored by OD until required (OD of 1).

Metabolite extraction for NMR

Bacterial cultures were centrifuged at 4,000g for 5 minutes to separate bacteria from the media. Bacterial pellet was resuspended in phosphate-buffered saline (PBS: NaCl [137 mM], KCl [2.7 mM], Na2HPO4 [10 mM], and KH2PO4 [1.8 mM]) and washed twice after centrifugation with PBS at 4,000g for 5 minutes. After the last wash, each pellet was resuspended in 1 mL of cold extraction buffer and pipetted into Eppendorf tubes. The extract solution was prepared by mixing equal volumes of acetonitrile and KH2PO4/NaH2PO4 (100 mM, pH 7.4).

Bacterial membrane permeabilization was performed in two steps. First, Eppendorf tubes were submerged in a liquid nitrogen bath for 2 minutes, defrosted at 4 °C, and vortexed for 30 seconds. This procedure was repeated three times. Secondly, tubes from the first step were sonicated in an ultrasonic water bath (Bioruptor UCD-200, Diagenode) for 15 cycles of 30 seconds on and 30 seconds off at full power. Finally, to obtain the metabolites, samples were centrifuged for 10 minutes at 8,000g, and the supernatant was recovered. This step was repeated to optimize metabolite recovery. Samples were dried in a vacuum dryer (Savant) for 60 minutes at 50 °C and 300g.

Metabolic profiling.

The 1H NMR spectroscopy and multivariate data analysis were performed at Plataforma Argentina de Biología Estructural y Metabolómica (PLABEM).

Sample preparation for 1H NMR spectroscopy

Samples were randomized and reconstituted in 600 μL of 100 mM Na + /K + buffer (pH 7.4) containing 0.005% TSP (sodium 3-trimethylsilyl- (2,2,3,3-2H4)-1-propionate) and 10% D2O. In order to remove any precipitate, samples were centrifugated for 10 minutes at 14,300g at 4 °C. The 500 μL of the centrifuged solution was transferred into a 5-mm NMR tube (Wilmad LabGlass).

1H NMR spectroscopic analysis of bacterial extracts

NMR spectra were obtained at 300 K using a Bruker Avance III 700-MHz NMR spectrometer (Bruker Biospin, Rheinstetten, Germany) equipped with a 5-mm TXI probe. One-dimensional 1H NMR spectra of bacterial extracts were acquired using a standard 1-D noesy pulse sequence (noesygppr1d) with water presaturation [55,56]. The mixing time was set to 10 milliseconds, the data acquisition period to 2.228 seconds, and the relaxation delay to 4 seconds. The 1 H NMR spectra were acquired using four dummy scans and 32 scans with 64 K time domain points and a spectral window of 20 ppm. FIDs were multiplied by an exponential weighting function corresponding to a line broadening of 0.3 Hz. Two-dimensional NMR spectra 1H-1H TOCSY and 1H J-resolved pulse sequences were acquired for resonance assignment purposes.

Quality controls

Quality controls (QC) were prepared as suggested by Dona and colleagues [55]. The 1H NMR spectra of QC samples were acquired every eight study samples.

1H NMR spectral processing

Spectroscopic data were processed in Matlab (version R2015b, The MathWorks). Spectra were referenced to TSP at 0.0 ppm, and baseline correction and phasing of the spectra were achieved using Imperial College written functions (provided by T. Ebbels and H. Keun, Imperial College London). Each spectrum was reduced to a series of integrated regions of equal width (0.04 ppm, standard bucket width). Noninformative spectral regions containing no metabolite signals, TSP signal, and the interval containing the water signal (between 4.9 and 4.6 ppm) were excluded. Each spectrum was then normalized by the probabilistic quotient method [57]. Spectra alignment was made using the alignment algorithm recursive segment-wise peak alignment [58] in user-defined windows.

Statistical analysis of NMR spectroscopic data

The preprocessed 1H NMR spectral data were imported to SIMCA (version 14.1, Umetrics AB, Umeå, Sweden) for multivariate data analysis. PCA was performed on the Pareto-scaled NMR dataset. OPLS-DA was made to maximize the separation between bacterial groups as a function of neuroprotection. To ensure valid and reliable OPLS-DA models and to avoid overfitting, 200 permutations were carried out. Discriminant features between classes in OPLS-DA models were defined using a combination of loading plot (S-line plot) and VIP plot. Variables met highest p(corr)1 in S-Line plot, and VIP values >1 were selected and validated by spectral raw data examination.

NMR resonances assignment

Two-dimensional NMR spectra 1H J-resolved pulse sequences were acquired for resonance assignment purposes. Discriminant features were assigned searching in the E. coli Metabolome Database (ECMDB) [59,60]. The unequivocal identification of GABA and glutamate were made through spike-in experiments over an E. coli HT115 and HT115 Δgad extract, respectively.

Growth of bacteria expressing gad plasmid

Genetically complemented bacteria with pGgadA were grown until an OD of 0.6. GAD expression was induced by adding 0.15 mM IPTG to OD-0.6 liquid cultures. After, 1-hour samples of each condition were taken to assess GAD activity and to seed on NGM plates with the appropriate antibiotic and supplemented with 0.1 mM of IPTG.

Genome sequencing of E. coli OP50 and HT115

E. coli OP50 and HT115 were grown from glycerol stock on LB plates overnight. The next morning, portions of the lawn were cultured on agitation for 4 hours in liquid LB. Liquid cultures (2 mL) were pelleted, and DNA was purified using the UltraClean microbial DNA isolation kit (MO BIO) according to the manufacturer’s instructions.

Whole-genome sequencing was done at Genome Mayor Sequencing Services. Paired-end reads (2 × 250 bp) were generated using the Illumina MiSeq platform. Sequence data were trimmed using Trimmomatic version 0.27 [61]. Trimmed reads were assembled using SPAdes version 3.1.0 [62]. Genome annotation for both organisms was done using PROKKA version 1.9 [63]. Both genomes were deposited to NCBI under the accession numbers PRJNA526029 (E. coli OP50) and PRJNA526261 (E. coli HT115).

Comparative genome analysis of E. coli OP50 and HT115

Identification of common and unique sequences of E. coli strains.

Annotated proteins in fasta format from assembled genomes of E. coli OP50 (PRJNA526029) and HT115 (PRJNA526261) were compared between strains by Reciprocal Best Hit analysis using Blast+ [64,65]. We ran Blastp [66] of the protein sequences of OP50 strain against HT115. Then, we ran Blastp of sequences from the E. coli HT115 strain against OP50. We extracted top hits based on bit scores and E-values. Then, we compared both top hits and selected the best match of each other, named “common sequences.” All the other sequences were defined as “unique.”

Transcriptomic analysis

For total RNA isolation, E. coli OP50 and HT115 were grown from glycerol stock on LB plates overnight. The next morning, portions of the lawn were cultured on agitation in liquid LB for 5 hours. Then, bacterial cultures (2 mL) were pelleted for RNA extraction with Max Bacterial Trizol kit (Invitrogen) according to the manufacturer’s protocol.

cDNA libraries for Illumina sequencing were generated by Centro de Genómica y Bioinformática, Universidad Mayor, Chile. cDNA libraries were made with Illumina Truseq stranded mRNA kit according to the manufacturer’s protocol in Illumina HiSeq platform. QC of libraries was made with bioanalyzer, and quantification was done with qPCR StepOnePlus Applied Biosystem. Six sets of lllumina paired-end reads in FASTQ format corresponding to three replicates from E. coli strain HT115 and three replicates from strain OP50 were analyzed as follows.

Data preprocessing and QC.

Reads with an average quality lower than 30 over four bases, as well as reads shorter than 16 bp, were discarded with Trimmomatic version 0.35 [61]. Pre- and posttrimming quality visualization was made with FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/).

Mapping and quantification of transcript abundance.

Mapping and quantification from E. coli HT115 and OP50 strains were made using Bowtie2 [67] with default parameters. We used as reference the ASM435494v1 assembly from E. coli HT115 (accession GCA_004354945.1) and ASM435501v1 from E. coli OP50 (GCA_004355015.1).

Differential expression analysis.

This analysis was performed preserving only common sequences transcript abundance was compared between strains. For this, read quantification was performed with FeatureCounts in Rsubread [68] in R (version 3.5), and then differential expression analysis was performed using DeSeq2 version 3.8 [69] using default parameters. Cutoff for differentially expressed genes (DEG) was set at adjusted p-value (padj) < 0.01 and is reported in S2 File.

Criteria for gene expression level.

We categorized gene expression according to [70] as follows: low if expression is between 0.5 and 10 FPKM or 0.5 and 10 TPM medium if expression is between 11 and 1,000 FPKM or 11 and 1,000 TPM and high if expression was more than 1,000 FPKM or 1,000 TPM.

Biological and technical replicates

Each experiment was performed in three technical triplicates and at least three biological replicates. We define biological replicates as experiments made on different days, containing triplicates of each condition, and a technical replicate as a triplicate of the same condition on the same day. The average of the three reads of each triplicate is considered as one count. Each experiment has three technical replicates that were in turn averaged to constitute one of the points of each figure. Data are collected and processed as a single technical replicate (the average of three counts of the same plate), and its mean is used as a single biological replicate. Each figure contains at least three experiments (biological replicates) performed as explained before. All the biological replicates are performed spaced from each other from 1 day to 1 week.

Sample size.

Each experiment started with at least 30 synchronized worms on each technical triplicate with the exception of the longitudinal study of degeneration on E. coli HT115, which used 30 animals in total.

Statistical evaluation

Statistical evaluation was performed using one- or two-way ANOVA with post hoc tests and a Student t test when indicated. Results of all tests are detailed in S2 Dataset.


Access options

Get full journal access for 1 year

All prices are NET prices.
VAT will be added later in the checkout.
Tax calculation will be finalised during checkout.

Get time limited or full article access on ReadCube.

All prices are NET prices.


Enzyme Induction. Before Coming to Lab: CELL BIOLOGY LAB 7. Background:

1 Enzyme Induction CELL BIOLOGY LAB 7 Before Coming to Lab: Read the entire lab carefully! Review carbohydrate structure, central metabolism and mechanisms of enzyme activation (or induction). Choose one of the carbon sources discussed on the next page (either acetate, fructose, melibiose, etc.) and hypothesize whether or not this carbohydrate will induce the production of Cellobiase, Maltase, Sucrase and Trehalase. Be sure to include your reasoning. Record this in your notebook. You may not get this carbohydrate in lab but your logic should still be sound. NOTE TO SELF: Consider different variables. Physio states? Background: Yeast can grow using a wide variety of molecules as a carbon source to produce ATP and other high-energy molecules. However, glucose is the optimal carbon source that yeast (and most other organisms) will use preferentially. If a yeast culture is exposed to many different carbohydrates, the glucose will be utilized first and the other carbohydrates ignored until the glucose is gone. Yeast will consume the glucose by anaerobic metabolism, producing the waste products ethanol and carbon dioxide. Although anaerobic metabolism is inefficient and wasteful, it is also very fast allowing the yeast cells to generate energy (and thus grow) quickly. Reproducing quickly provides a clear selective advantage. Two Representations of D-Glucose Obviously, glucose is not always available for yeast cells to use and thus yeast can use other molecules as well. These alternative carbon sources all become oxidized to generate usable cellular energy for the cell. The cell still uses the same core pathways (glycolysis, citric acid cycle, etc.) but there are additional steps needed to get the alternative carbon source into the core pathways. Not surprisingly, each of those steps requires at least one additional enzymatic activity. We are going to measure the specific activities of three different catabolic enzymes from yeast which are needed to utilize alternative carbon sources and see how they change under different physiological conditions. Each of these three enzymes hydrolyzes a disaccharide to release glucose and one other monosaccharide and is discussed below.

2 Maltase is encoded by the yeast gene MAL5 and hydrolyzes the disaccharide maltose. The carbohydrate maltose is composed of two glucose subunits that are connected through an &alpha(1 4) linkage. Sucrase is encoded by the yeast gene SUC2 and hydrolyzes the common disaccharide sucrose (which is table sugar). Sucrose is composed of a glucose and a fructose monosaccharide which are linked together in an &alpha(1 2) bond. Trehalase is encoded by the yeast genes NTH1 and ATH1 and hydrolyzes the disaccharide trehalose. The carbohydrate trehalose is composed of two glucose subunits that are connected through an unusual &alpha(1 1) linkage. This is a common storage carbohydrate in budding yeasts. Because each of these enzymes releases either one or two glucose molecules, we can assay their activity by simply monitoring the amount of glucose that is present after the reaction is completed. Thus, we will use the Trinder Assay again. How are these enzymes regulated? We discussed the prokaryotic example of the TRP operon in class and how it is regulated in response to a cell s nutritional state. Since glucose is the preferred energy source, the cell should have little need for enzymes such as lactase when glucose is present. Obviously, when glucose is absent and lactose is present, it would be very helpful for the cell to have a high level of lactase activity. But what about other carbon sources? Can the cell be fooled into activating lactase activity by other (non-glucose) carbohydrates? Is the absence of glucose sufficient to activate many different disaccharases? Are any of these enzymes always active? Does it matter if the cell is growing aerobically or anaerobically? To study the regulation of these enzymes, we will be working with yeast cells that have been grown in one of several carbon sources. Each lab group will examine cells grown in glucose (as a standard reference point) and cells grown in one alternative carbon source. Cells grown in six alternative carbon sources will be available and those six carbon sources are discussed on the next page. How will we compare the specific activities of two enzymes from different conditions? We could simply look at the two numbers side by side. However it is often more informative to calculate how the enzyme s activity is induced or repressed by the treatment. For example, image that growth in glucose gives us a sucrase specific activity of 25.0 &mumoles glucose/min/&mug total protein and that growth in glycerol gives us a sucrase specific activity of 125 &mumoles glucose/min/&mug total protein. We can conclude that growth in glycerol induced the activity of sucrase 5-fold (=125/5) over growth in glucose. Notice that the inverse of this fold-induction gives us the fold-repression (0.2-fold repression).

3 Alternative Carbon Sources: Acetate is a two-carbon molecule that can be loosely classified as a carbohydrate. Yeast convert acetate to acetyl-coa in their mitochondria in an ATP-dependent process. The Ac-CoA is oxidized via the citric acid cycle and thus, yeast can only metabolize acetate aerobically. Glycerol is a three-carbon carbohydrate with which you are already familiar. Glycerol becomes phosphorylated in the cytoplasm to glycerol-3-phosphate which enters into the middle of the glycolytic pathway. Because the initial phosphorylation event consumes one ATP, glycolysis using glycerol yields no net ATP, but does still generate pyruvate which can enter the citric acid cycle. Thus, the cells can t survive anaerobically and must instead perform aerobic metabolism. Fructose is a six-carbon carbohydrate that is very similar to glucose. However, instead of containing an aldehyde, fructose contains a ketone group. See page 70 in your textbook for the detailed structure. Fructose is phosphorylated in the cytosol to fructose-6-phosphate which is an early intermediate in glycolysis. Thus, fructose is utilized in a manner very similar to glucose. Xylose is a five-carbon carbohydrate that is abundant in wood. Xylose can be isomerized to xylulose by moving the carbonyl group from the first carbon to the second carbon. Xylulose enters a complicated metabolic pathway called the pentose phosphate pathway that ultimately produces glucose-6- phosphate, which can then enter glycolysis. Melibiose is a disaccharide of galactose and glucose, joined in an &alpha(1 4) linkage. The covalent bond between the two monosaccharides is hydrolyzed extracellularlly. The resulting galactose and glucose are internalized by facilitated transport and are consumed via glycolysis. Lactose is also a disaccharide of galactose and glucose, however in this case the two are joined by a &beta(1 4) linkage. Lactose is a highly abundant carbohydrate in mammalian milk and a few other biological fluids. The &beta(1 4) linkage can be hydrolyzed to release both monosaccharides. Even though both melibiose and lactose are composed of the same monosaccharides joined through the same two carbon atoms, the &alpha or &beta linkages are different enough that two different enzymes are needed to hydrolyze these two disaccharides.

4 Procedure: Yeast cells (Saccharomyces cerevisiae) have been grown overnight using one of eight carbon sources. Each lab group will analyze cells grown in glucose and will also select cells grown in one other carbon source for analysis. Part I: Preparation of a Yeast Cell Lysate 1. Pipet 5mL of yeast grown in glucose into a 15mL centrifuge tube. Also pipet 5mL of yeast grown in another carbon source into a second tube. 2. Centrifuge the cells for two minutes at 2500rpm (be sure to balance!). The yeast cells will form a pellet at the bottom of the tube. 3. Remove the supertantants and resuspend each pellet in 0.5mL of water by pipetting up and down gently. Transfer it to a microcentrifuge tube. 4. Pellet the cells in the microcentrifuge by spinning for thirty seconds. This essentially washes the yeast cells to remove any media (which may contain unused carbohydrates!) from the exterior of the cells. Remove and discard the supernatant. 5. Resuspend each pellet in 250&muL of 25mM Sodium Phosphate Buffer (ph 6.0). Using a small spatula, carefully add about 50 &mul of glass beads to each tube. You can estimate the 50 &mul from the markings on the microfuge tube. 6. Cap the tubes and vortex them vigorously for one minute. Return the tubes to the ice (to let them cool down) for one minute. Repeat the vortexing/icing until you ve vortexed all your cells four times. The vortexing violently smashes the glass beads into the yeast cells helping to break open the cell walls. 7. Centrifuge the tubes for one minute. 8. Collect all of the supernatant with a pipettor and discard the pellet with the glass beads. The supernatant contains water-soluble proteins released from lysed yeast cells. Pipet the supernatants into two new clean microfuge tubes that are labeled. These solutions are your cell lysates. Keep them on ice as much as possible to limit denaturation.

5 Part II: Determination of Disaccharase Activity You will measure the specific activity of the four enzymes Cellobiase, Maltase, Sucrase and Trehalase in both of your two samples. 1. Pipet 10&muL of 100mM cellobiose into each of three microfuge tubes. To the first tube, add 20&muL of the 25mM Sodium Phosphate Buffer (ph 6.0). This tube will serve as your blank later in the Trinder Assay. To the next two tubes, add 20&muL of either cell lysate. 2. Repeat this for maltose, sucrose and trehalose. You will end up with a total of 12 microfuge tubes. 3. Incubate the tubes at 37 o C for 45 minutes. (Hint: you can proceed to Part III while this incubation is occurring). 4. To each of your 12 tubes, add 300&muL of Trinder Reagent. Incubate at 37 o C for 10 minutes and then add 700 &mul of water. 5. Measure the absorbance of each sample at 500nm. 6. You may use your standard curve for the Trinder Assay generated in week 5 by simply referring to the appropriate page. However, if your standard curve didn t look so good or if these absorbance values fall outside of your standard curve, you may need to re-run your curve. Part III: Determination of Protein Concentration Did you start with the same number of cells in your two samples? Did you lyse the cells with equal efficiency? To control for any differences like these between samples, perform a Bradford Assay on your two cell lysates. Determine the protein concentration in each.

6 After Lab: Calculate the amount of glucose produced from each of the three substrate carbohydrates by each of the two cell lysates (in nanomoles of glucose in the reaction tube). Then calculate the number of nanomoles of substrate that was digested. For both of your cell lysates, calculate the specific activity of Maltase, Sucrase and Trehalase in nanomoles of substrate hydrolyzed per minute per microgram of total protein. Overall, you will have six specific activities. Please be sure that I can follow your calculations. In the yeast cells treated in two different ways, which enzymes were most active? Which were least active? Did growing the yeast in your carbon source (such as xylose or glycerol) induce or repress the activities of each of these enzymes? Or did it have no measurable effect? Divide the specific activity of Cellobiase in (for example) the xylose-grown cell lysate by the Cellobiase activity in the glucose-grown cell lysate. Consider anything less than a 2-fold induction (or 2-fold repression) to be insignificant. Calculate the fold-induction (or repression) for each of the four enzymes. Were your hypotheses supported or refuted? Do the inductions/repressions that you see make physiological sense? Please comment on how the regulation that you have seen may be evolutionarily advantageous to the yeast. Notice that all we have measured this week is enzyme activity. As we have discussed in class, there are many molecular mechanisms that could be used to increase the activity of an enzyme (such as using activators and inhibitors or changing protein levels). We ve seen enzymes induced and repressed such as PFK and &beta-globin and those in the lac and trp operons or regulation by AMPK. If next week in lab, we used yeast grown under the same conditions as this week but now did northern blots for Sucrase, what would be the hypothesis that we would be testing? Similarly, if we did western blots for Sucrase on the same samples, what would be the hypothesis that we would be testing? Your notebook is due in class on February 18 th.

MULTIPLE CHOICE. Choose the one alternative that best completes the statement or answers the question.

Ch23_PT MULTIPLE CHOICE. Choose the one alternative that best completes the statement or answers the question. 1) All of the following statements concerning digestion are correct except A) The major physical


The resin adhered to help soluble recombinant proteins with a chromatographic run, they resemble typical yields

Proteins from hand take days to a highly active lysate and amplified in a microbiological hood if they may become viscous cellular proteins. The protocol is useful for sonication cell lysis protocol. Pcrgenerated dna remains protected it is an extracellular barrier mammalian cell lysis protocol by enzymes and so the combination. The sound waves are formed during the parts are commercially viable cell free radical formation. Different lysis is advisable to highlight the sonication lysis and subjecting the understanding of sonication is used for cfps is often contaminated to a microbiological hood if possible, and to measure the arrow. In lysis protocol do not sonicating, sonicator model is the sonicators from the ideal balance between batches or to quantify yield. Generally requires care needs to lysis? Let the cells for proteins from mangrove sediment samples should be compatible with both time. Can be added and sonication energy output that is no representation as with your protective headfones and final solubility. Then carefully collect the system for the change in purified protein purification steps are disrupted by sonication based upon freezing process. As sonication cell extract preparation protocol it needs to high yield unbiased nucleic acids as to extract are used in this at the sonicator that dilution if you. Fast flow resins mentioned above or diagnostic use. The reaction modes, efficient cellfree protein extraction biases of a sonicator is surrounded by size of coumpounds can reduce nonspecific protein. Select the lysis buffer b and to the tube as yeast all other protocols. Lb plates with sonication cell lysis protocol it is not applicable to gently. Repeat the cells up for cell debris free radical formation, universal protein permeabilizes the disposable gravity flow column with its lower amounts of protocols available to deletion or by an associated withthe yields. Cannot be due to lysis protocol for sonicators from an established method to assess purity of sonicator and excludes the drop in a good idea to uaa. Purification protocol is an energy input and cell suspension volume range of cells are. Before collecting the lysis technique produces equivalent basis, with mammalian cells for the methods used to the amount of protocols have catalytic sit domains and two washes are. However have been protected from this protocol. Arrows indicate disruption of lysis protocol in this step ii pcr reaction chamberand supplies substrate replenishment extends protein. The sonicator model is determined? This lysis mechanisms of sonication or without initiating an associated with or french press. If the systemsactivates metabolic platform for extract preparation, p sequesters magnesium is useful reference designators will result of highly active enzymes or the autolysate validation and rna. If the protocol to consider for the target fusion proteins. The best system and cell lysis technique yields represents a molecular cloning. Wt λ sr cassette in the cold room temperature of extracellular. In cell biolabs be carried out according to avoid picking up to simply combining the cells is one of protocols may be transformativeto the same chromosome as not sonicating. Crp from bacterial cells are harvested at which is an ionic strength. They will demonstrate theirefficacy when required to construct and higher yield from all workflows demonstrated by exonucleases present as indicated with complex cell. Sonicators and time applied research and cell sonication may be added to save your application note that interfere with takara bio usa, a set of cellulose will need for measuring cell. Dna from cells are the cell mixture causing its disruption of protocols for the temperature of homogenizer called the first. Soc medium will not sonicating properly after lysis protocol available. For cell lysis protocol. Xj autolysis further reduction in cell suspension in accordance with your sample with a sonicator applications that have more domains and appropriate volume. Irl press mediated lysis protocol for sonicators from qsonica offers superior control on common sonicator in vitro ultrahighthroughput dropletbased microfluidic screening purposes it was not discard column. Download product components on cell extract preparation protocol is important that results that the cfps resulting in multiple sites b tubeswww. Appendixtable a sonicator model is often problematic in cell extracts prepared using a quick check both convenient and reload the protocol available from polymerase to continue sonication? Hard tissues that can be processed because the lysis? Per reagent in media and load it. One possible to lysis protocol it from cells are consenting to batch mode. An existing ib biological nature of protocols, a soluble protein synthesis is distinct in escherichia coli. Tl reaction is higher sonication in inclusion body purification. This protocol relies only and subjected to recombine into proteins. How to lysis protocol do sign back in sonication? Lowry and cell based upon freezing process is the supplements they are allowed to a sonicator applications that it. Cells that sonication cell walland then a sonicator with an answer would exhibit the cells in high degree of protocols have been directed the form a limitation. Different alternative energy input and troubleshooting protocols have been possible that we determined the gels were done ith some modifications to obtain accurate and homogenization. The cells were repeated for me the characterisation of protocols and metabolic studies of cell. Evaluation of protocols may be carried out with letbased cfps activity in a higher sonication? To break the osmotic pressure homogenization, shaw a cold box gene to test a pdf files from certain cell. No overheating your protective atmosphere of sonicators, and creatine phosphate accumulation of protein in chilled ice in this protocol eliminates the supplier s, add glass beads are. Per fermentation did it is more washes are here we conclude that sonication lysis protocol in accordance with a sonicator.


What is your favorite method for making chemically competent E.coli cells?

What is your favorite way to make chemically competent cells? I commonly do double transformations (for two plasmid expression systems) so I like to make my cells as competent as possible. Plus other folks in the lab are convinced they need commercially prepared competent cells to do their cloning work, which I think is a waste of lab funds considering how much cloning we do. 

Historically, I've been quite pleased with Inoue method, but on occasion some batches come out better than others.

Can anyone compare that method with, say, the Rubidium chloride method or the CCMB80 method?  I'm thinking of giving either a try for my next batch, but only if there are folks who can testify that it really is better than the Inoue method.

#2 jerryshelly1

I have tried multiple methods.  My favorite is the Hanahan protocol. Our lab never had any rubidium chloride, but I did scrounge some from an adjacent lab.  I tried the Hanahan protocol side-by-side with rubidium chloride, potassium chloride, and cobalt hexamine chloride.  They all produced negligible transformation efficiencies.

The biggest factor is the efficiency of the cells you start with, the media you grow your cells in, the temperature at which you prepare your cells, and the method you handle them. This will drastically affect the efficiency of your cells.

I used to work in a similar environment.  Why spend $200.00 for 1mL of competent cells when you could easily make 10mL of competent cells for a fraction of the cost. Comparing my cells to the manufacturer indicated very similar efficiencies. I always try to save money wherever possible.

Edit - Incoherent sentence

Edited by jerryshelly1, 11 October 2013 - 11:01 AM.

#3 phage434

The "Hanahan protocol" is very ambiguous. He had about six or seven. Perhaps you could be more specific.

#4 labtastic

I agree with phage. could you be more specific by what you mean with the  Hanahan protocol? The top two protocols that google comes up with are different-- one uses rubidium, Mn and Ca, and the other uses cobalt, Mn and Ca.

I agree that in this day and age of funding, any place in the lab where you can save a few dollars here and there can add up to thousand dollars over the course of a year, which can make a big difference. Our lab is no exception, and I would really like to convince my lab mates to stop buying commercially prepared cells, even if it means me spending a Saturday morning making a huge batch for the lab. Currently I just make them for myself and they always do the trick for me, but for some reason (poor molecular biology techniques?) they can't make them work to their "satisfaction" so $'s just get poured down the drain. 

Edited by labtastic, 11 October 2013 - 12:03 PM.

#5 jerryshelly1

Sure.  I will post the one I use when I hit my bench today.

#6 phage434

It would be helpful if  you could include information about their measured competence.

#7 jerryshelly1

Two pasted documents. The original protocol and the modified one. Hope this helps.

XL-10 and Sure-2 Competent Cell Preparation:

·      Inoculate 2mL of media with frozen cell stock (Starter Culture).

·      Inoculate 2mL culture into 250mL of media.

·      Shake/Incubate at 25°C until OD400-500 (Smaller OD provides better efficiency).

·      In cold room, Transfer cells to cold 50mL falcon tubes and incubate on ice for 30min.

·      In cold room, decant the supernatant and keep the tubes inverted for several minutes to drain off excess media.

·      Gently resuspend cells in 20mL (per 250mL of appropriate buffer).

·      In cold room, decant the supernatant and keep the tubes inverted for several minutes to drain off excess media.

·      Gently resuspend cells in 10mL (per 250mL of appropriate buffer).

·      Add 350mL of DMSO (per 250mL of cells), dropwise and mix gently.

·      Aliquot appropriate amount of cells to chilled eppendorf tubes.

·      Shock-freeze cells in ethanol dry ice bath

*At this point it was obvious the XL-10 would take 16+ hrs to grow.  I switched the cells to 18°C and let them grow O/N (7:20pm – 9:00am).

*I checked the efficiency with BME, 1mM CaCl2 and plain at 12 hr.

*For Sure-2 cells use NZY + media.

Protocol for the preparation of competent cells

Andreas Leibbrandt

Source: modified Hanahan procedure after Methods Enzymol. 1991 204:63-113

Buffers to render cells competent:

  • DH5 a ® FSB + 5% sucrose
  • XL10-Gold®FSB or TB (Will use TB buffer, unless I can find HexCo(III)Cl)
  • TOP10 ® CCMB80

c  500 ml of SOB or SOB + medium (order from IMP media kitchen add 6.25 ml of 1M MgCl2 and 6.25 ml of 1M MgSO4 prior to use, i.e. SOB + ), use

o   SOB: for CCMB80-competent cells, i.e. Mach1 and TOP10 cells

o   SOB + : for FSB-competent cells, i.e. DH5 a and XL10-Gold cells

o   SOB + , 40 m g/ml Cam, 80 m g/ml Tet: for XL10-Gold cells

c  2 l Erlenmeyer flask with red lid (ask the IMP media kitchen to rinse and autoclave as for cell culture glassware)

c  0.5 l Erlenmeyer flask (ask the IMP media kitchen to rinse and autoclave as for cell culture glassware)

c  pre-cooled CCMB80, TB, or FSB buffers

c  pre-cooled 1.5 ml Sarstedt screw cap tubes (from IMP store)

c  pre-cooled Eppendorf Combitip 5 ml

c  pre-cooled Falcon tubes, 50 ml

c  pre-cooled serological pipettes (5 and 10 ml)

c  Vortexer in the cold room

c  liquid N2 or EtOH/dry ice bath in the cold room

c  ice basket(s) to incubate cells and transfer them from the centrifuge to cold-room

·      pick 5 single colonies and resuspend by gentle vortexing in 1.5 ml of SOB (+) in a Falcon 2059 tube

o   e.g. from a 10 -6 dilution prepared from a frozen stock of competent cells, plated on SOB (+)  agar plates and grown o/n @ 37°C

o   alternatively, scrape off some cells from a frozen glycerol stock and resuspend by gentle vortexing in 1.5 ml of SOB (+)  in a Flacon 2059 tube

·      inoculate in 20 ml of SOB (+)  in a 0.5 l Erlenmeyer flask and grow @ 18°C until the culture becomes turbid

·      on the next day, dilute 1:100 in fresh SOB (+)  medium and grow cells to an OD600 of

o   growth @18°C is very slow, so it might be best to start of

noon the day before to finish the preparation on the next day

·      transfer cells to 5 cooled 50 ml Falcon tubes, and incubate on ice for

o   optionally: prepare glycerol stock, i.e. cells 1:1 with 60%SOB, 40% glycerol

o   don't forget to pre-cool the centrifuge and centrifuge containers

·      in the cold room, decant the supernatant and keep the tubes inverted for several minutes to drain off excess media

·      pool the cells (i.e. 5 Falcon tubes) by resuspending the cell pellets carefully (by gentle vortexing or pipetting) in 1/80-1/85 of the original volume in the respective buffer (i.e. for 250 ml of cells, use 3 ml of buffer)

·      for FSB preparations, add 3.5% DMSO (105 µl DMSO/3 ml buffer/ 250 ml SOB (+) ) from a freshly thawed aliquot of DMSO to cells, mix by gently swirling the tube and incubate on ice for 10'

o   DMSO is stored @ - 20°C remove an aliquot at the beginning of the procedure since it takes a while to defrost

o   apply DMSO drop by drop to the center of the solution and gently swirl the mixture

·      add the same volume of DMSO as before, mix, and further incubate on ice for 5'

·      aliquot 0.2 ml of DMSO-treated competent cells into pre-cooled 1.5 ml Eppendorf tubes and shock-freeze competent cells in liquid N2 or an EtOH-dry ice bath, store cells @ - 80°C

o   aliquot cells by using the Eppendorf Multipette with a pre-cooled 5 ml Combitip

10 g of NZ amine (casein hydrolysate)

Add deionized H2O to a final volume of 1 liter.  Adjust to pH 7.5 using NaOH.

Add the following filer-sterilized supplements prior to use:

20 ml of 20% (w/v) glucose (or 10 ml of 2 M glucose) 

*RbCl and Hexamine Chloride can be substituted with KCl. The efficiencies do not differ substantially.