This is the method I use routinely.

Procedure Remove tissue or fetus and fix in fresh 4% paraformaldehye/PBS (pH 7.0-7.5) for 1 hour at 4oC. Rinse three times, for thirty minutes each, with rinse buffer (recipe given below) at room temperature. Stain between 4 and 48 hours, typically overnight. About 90% of potential staining will occur in the first 24 hours. Post-fix overnight in 10% formalin (10% formalin is fine because freshness isn't critical) at 4oC. Recipes

• PBS:
  • 100 mM sodium phosphate (pH 7.0-7.5)
  • 150 mM NaCl
• Rinse buffer
  • 100 mM sodium phosphate (pH 7.3)
  • 2 mM MgCl2
  • 0.01% sodium deoxycholate
  • 0.02% NP-40 (by volume)
  • Stable at room temperature for at least a year
• Staining solution
  • Rinse buffer plus:
  • 5 mM potassium ferricyanide, from 0.5 M stock stored at room temperature in the dark
  • 5 mM potassium ferrocyanide, from 0.5 M stock stored at room temperature in the dark
  • 1 mg/ml X-gal, from 25 mg/ml stock in dimethylformamide (DMF), stored at -20oC
  • I almost always prepare it fresh from stock solutions.

Notes

• I typically transfer the post-fixed tissue to 70% ethanol for long term storage (up to several years). The stain is stable and ethanol whitens the tissue for better contrast in whole mount photos. Long term storage (longer than a couple weeks) in formalin can harden the tissue, making it more difficult to cut nice sections later.
• Increasing the pH will decrease the endogenous enzyme staining, at the expense of slightly weaker bacterial enzyme staining. Supposedly, about pH 8.5 is the most you can get away with (I've never gone above 8.0).
• You'll sometimes get crystal deposition on your tissue; this is unreacted X-gal. You can minimize it by using a 25 mg/ml X-gal stock (i.e. higher concentration of DMF in the staining solution), by using DMF and not DMSO (dimethylsulfoxide, which some people prefer because it's less toxic), and by pre-warming your sample and stain to 37o C(this only helps a little). In any case, X-gal crystals will dissolve after a couple days in ethanol or (more slowly) in PBS.
• The X-gal stock will slowly go yellow, but I keep using it until it gets intimidatingly bright. This happens much more slowly to stocks prepared in DMF than in DMSO.
• You can reuse the staining solution by filtering it and storing it at 4oC, though I've never stored it longer than a week or reused it more than once.
• You can increase the concentration of X-gal up to 4 mg/ml for significantly increased staining (and significantly increased X-gal budgets). Use the same concentration of X-gal stock solution so more of the DMF ends up in the stain, too.
• Note that potassium ferrocyanide is usually supplied as a trihydrate, at 422 dal. Potassium ferrocyanide is 329 dal.
• You may get inconsistent results using 10% formalin solutions (Raj Kapur). Some sources of confusion: Formaldehyde is a gas. The formaldehyde solution you can buy is called 100% formalin. Paraformaldehyde is polymerized formaldehyde that produces formaldehyde gas when dissolved in water. A 10% formalin solution should be equivalent to a 4% paraformaldehyde (by weight) solution. Whew! In the past I mostly made fresh 4% paraformaldehyde, but lately have gotten good result with frozen stocks of 20% paraformaldehyde in PBS, freshly diluted into additional PBS for fixation.
• I use the same procedure for slide-mounted frozen sections, fixed either before or after sectioning. Fixing first is usually easier. Tissue histology is somewhat better if you replace the deoxycholate/NP40 with 0.1% Triton X-100 (or omit detergents completely if your staining is strong enough without them).
• Beware of halting the staining early! You may see strong staining in some places or cells after just 4 hours and apparently none elsewhere. However, X-gal stain precipitation shows a strong threshold effect, with rapid accumulation of precipitate once the threshold is reached. You might find that four hours later an additional set of strongly staining cells has appeared. Unless background or stain bleeding is a significant problem, you're better off letting the reaction go for at least 12 hours. Nearly all sites that will stain have appeared by then, though extending staining to 24 hours will intensify the weaker ones somewhat.
• I find that fixing at room temperature with 4% paraformaldehye gives significantly less staining than when it's performed at 4 oC. I've been told this isn't the case for gluteraldehye fixation, though.
• In electron micrographs of X-gal stained tissues, the precipitate appears as tiny elongated crystals. If the crystals have started to dissolve during dehydration (i.e. too long in xylene), they'll be fuzzy instead.
• If you prepare your stain in lab glassware, be aware that some sorts of detergent contamination on lab glassware will cause the X-gal to precipitate out as a fine white cloud as soon as you add the X-gal stock to the bulk stain solution. It won't go back into solution with heating, so don't bother trying. Throw it out and prepare the stain in disposable plastic or carefully washed glassware.
• It's not unusual to see good staining in whole mount that is not too impressive after sectioning. Your options are: use thicker sections (vibratome sections are useful), or use darkfield optics. Under darkfield optics of thin (<20 mm) sections, the crystals appear bright pink and this is VERY sensitive; unfortunately, some people have reported their sections look fine in bright field but nothing appears under darkfield optics! I don't know why this happens to some people, but suspect they're leaving the section in xylene too long, and the crystals are starting to dissolve and lose the shininess that makes them stand out under darkfield optics.
• Some people will tell you it isn't possible to paraffin embed X-gal stained tissues because the xylene dissolves the stain. They're leaving the tissue or section in xylene too long.
• Increasing the K-ferri/ferro to 30 mM alters the subcellular localization of stain deposition as judged by electron micrographs (Josh Sanes).
• Stain penetration is highly variable between tissues. Whole E11.5 mice show good stain penetration, but it's somewhat less at E12.5, and by E13.5 it's clearly limited (cut them transversely at cervical or abdominal levels, or both, for better results).
• Adult brains won't stain well unless you cut them up. Good stain penetration is limited to about 250-500 mm (from the cut face or nearest ventricle). Vibratome sections of 50-200 mm give great results, but an excellent alternative for rapidly making sections from many samples is a metal block designed for crude slices (1 mm thick). I use the Rodent Brain Matrix (RBM-2000) from ASI Instruments [Warren, MI, U.S. , (800) 531-1105], in either saggital or coronal formats, to generate 1-2 mm sections for staining. You can later cut thin sections off the face of the 1-2 mm sections that stain. Sometimes the 1 mm sections curl a bit during fixation. For perfectly flat ones to be used for thin sectioning later, either use 2 mm crude sections, or first perfuse the animal with fix, quickly remove the brain into cold fix for 15 minutes, section as 1 mm slices, and fix an additional 45 minutes.
• Combining beta-galactosidase staining with in situ hybridization is not difficult. Do the X-gal staining first as usual (use cold 4% paraformaldehyde; using gluteraldehyde at this step my not be compatable with the later in situ hybridization). You'll probably want to prepare the solutions in DEPC-treated water, and consider extending the fix a bit if the tissue is thick (to better inactivate endogenous RNAses). Post-fix with your usual in situ hybridization fixation, and continue from there.
• Buying X-gal from the usual laboratory chemistry suppliers is a bad idea, since you can get it much more cheaply from alternative sources. The best I've seen are: Gold Biochemical $80 (US) per gram, (800) 248-7609 (U.S.). BioSynth International $46 (US) per gram, with a minimum 10 g order (find a friend!); (800) 270-2436 (U.S.), BioSynth AG (071) 43 01 90 (Switzerland).

Tissues with background staining

• Choroid plexus (faint) Lacrimal gland Submandibular gland Neonatal bone (faint) Ovary Mucosal layer of stomach, small and large intestines (intense!) Endolymphatic diverticular appendage of the developing otic vesicle (mouse E11.5-13.5) (faint)

Some Variations

• Do the first wash at 4oC instead of room temperature (I commonly do this with thicker samples)
• Agitate samples during washing and/or staining (I do this sometimes for larger sized tissues, >100-200 mg).
• Increase the detergent concentrations to 0.1 % sodium deoxycholate, and 0.2 % NP-40 temperature (from Joyner/Rossant/Soriano labs). I don't know if this reduces the quality of the the histology.
• Fix in 0.2% gluteraldehyde, 2% formaldehyde, 0.1 M phosphate (pH 7.3), 5 mM EGTA (not EDTA!), 2 mM MgCl2 for 1 hour at room temperature (from Joyner/Rossant/Soriano labs).
• There is a report with some interesting suggestions for improving bacterial beta-galactosidase activity and decreasing the endogenous mammalian enzymes (Young, D.C. et al, Analytical Biochemistry 215:24-30, 1993). This work was done with in vitro solution assays, though. I have NOT tried it yet with histological staining, but plan to some time. The best results were reported with a 1 hour heat pretreatment at 50oC (no benefit at 45o, and 55o decreases the bacterial activity also), and a modified enzyme buffer:
  • 100 mM HEPES
  • 5 mM DTT
  • 1 mM MgSO4
  • 2% Triton X-100
  The optimum pH for best histochemisty is likely to be around 7.5 (7.0-8.0). I plan to do the rinses in the described buffer, replace my third rinse with one for 1 hour at 50oC, and then use a stain solution of the rinse plus K-ferri/ferro and X-gal at the usual concentrations.
• BioSynth (see above for contact), also sells two variant beta-galactosidase stains, magenta-gal and salmon-gal, named for the colors of the precipitates they produce. I've used the magenta-gal and found it to be as sensitive as X-gal, but I saw more deposition of precipitate outside the cells that actually expressed the lacZ gene. Next time I use it, I plan to increase the K-ferri/ferro concentrations to 30 mM each. This stain should give terrific color contrast in combinationwith in situ hybridization using NBT/BCIP stain.

---

Clearing X-gal Stained EmbryosIf you've never done this, you'll be delighted with glassy pale yellow embryos with intensely blue stained tissues inside. They look like jewelry.

 

Methyl SalicylateThis is the method I use. Post-fix embryos in 10% formalin overnight at 4oC. Wash with distilled water twice, each for 30 minutes at room temperature. Dehydrate as follows (all at room temperature with agitation): 30 minutes in 70% ethanol, 30 minutes in 95% ethanol, two times in 100% ethanol for 30 minutes each. Transfer to 100% methyl salicylate, and agitate at room temperature (or 4oC), until the tissue clears. Clearing takes about 15 minutes for E10.5 mice, and about an hour for E15.5 fetuses.Notes

• I've never tested how large a sample you can clear effectively. The key is to eliminate all traces of water, else you'll get a white precipitate when you add the methyl salicylate.
• The X-gal stain will slowly dissolve in the methyl salicylate, decreasing slightly overnight, and generally is completly gone within four days.
• Methyl salicylate dissolves polystyrene!! Glass or polyethylene is fine.
• Methyl salicylate isn't especially toxic, and leaves your lab smelling like a peppermint Lifesavers candy factory.

Cedar Wood Oil Dehydrate embryos through graded ethanol steps up to 100% ethanol Place embryos in Xylene:100% ethanol (1:1) for 1 hour. Transfer embryos to high quality cedarwood oil and rock gently, 1-4 hours. Replace cedarwood oil once and rock gently overnight.Notes

• Originally from Bob Hammer; I haven't tried this.
• The X-gal stain is apparently stable in cedarwood oil.
• Not all sources of cedarwood oil will work. One that does is from Polyscience, catalog #0485, but I hear they won't be carrying it for much longer.
• Cedarwood oil isn't toxic, and leaves your lab smelling like a nice cedar chest

Benzyl Benzoate Dehydrate embryos through graded steps into 100% methanol. Transfer embryos into benzyl benzoate:benzyl alcohol (1:1).Notes

• I haven't used this.
• Benzyl benzoate is toxic, and doesn't smell nice either.

---

X-Gal Histochemisty of Culture CellsThis procedure is faster and simpler than the procedure for whole tissues. An important difference is the addition of gluteraldehyde to the fix, to more firmly attach cells to the culture plate.

Reference: EMBO 5:3133 (1986), modified by Jaques Peschon.

 

Aspirate off medium and rinse once with PBS (pH 7.3) Fix 5 minutes on ice. Rinse three times with PBS (1-4 minutes each). Stain at 37oC, overnight. Rinse once with PBS, and store in PBS at 4oC

Recipes

• Fixative
  • 2.7 ml of 37% formaldehyde (formalin)
  • 0.4 ml gluteraldehyde (Sigma sells 25% stock solutions)
  • to 50 ml with cold PBS (pH 7.3)
• Stain
  • 1.0 ml X-gal stock (25 mg/ml in DMF)
  • 50 ul MgCl2
  • 41 mg K-ferricyanide
  • 53 mg K-ferrocyanide
  • to 25 ml with PBS (pH 7.3)

Notes

• Don't do the staining in the tissue culture incubator! They typically use a 5% CO2 atmosphere, which causes the pH of the stain solution to drop, and you'll probably have background problems.