ß-galactosidase assays in yeast

This page is intended to supplement our review in Analytical Biochemistry

There is NO such thing as STANDARD procedure for ß-gal in yeast. There are dozens of techniques, each with its advantages and drawbacks.

	What are the chances of identifying a ß-gal assay protocol described in a paper? To answer this question, we screened 80 on-line papers from 1999, which described determination of ß-gal activity in yeast (mainly from PNAS, J Biol Chem, Cell/Molecular Cell and the Nature family; list is available upon request). As Figure 1 shows, about 50% of papers did a good work either by providing a protocol directly as part of Materials and Methods section (shown as "0" on the graph), or referring to an actual source (shown as "1" ). For a significant fraction of articles, however, getting to an actual protocol was only possible following lengthy chains of 2, 3, 4 or more references (mainly including self-quotations of the same authors). Finally, about 20% of papers provided just wrong or non-existing references.
How old are sources of these protocols? Analysis of the same set of papers is shown in Figure 2, where protocols, to which references are made, are arranged by year.
	What are these protocols? Only a fraction of references points out to the original research paper. Typically, the sources are either specialized protocol books, or companies' protocols. CP: Current Protocols; Bartel: combined Bartel, P. L. & Fields, S. (1995) Methods Enzymol. , AND  Fields, S., Sternglanz, R., Chien, C. T., and Bartel, P. L. (1993) in Cellular Interactions in Development: A Practical Approach; Breeden:Breeden, L. & Nasmyth, K. (1985) Cold Spring Symp. Quant. Biol. 50, 643-650; Meth. yeast: Methods in Yeast Genetics, combined editions 1990, 1994, 1997; Miller: J. H. Miller, Experiments in Molecular Genetics (1972).

What is the origin of the galactosidase gene assayed in yeast?
S. cerevisiae yeast (at least some strains ) have their own galactosidase  ( a-galactosidase, also known as melibiase (MEL1), a-D-galactoside galactohydrolase, EC 3.2.1.22). Although there are now reporter systems based on the MEL1 gene, typically the E.coli lacZ gene is used, encoding ß-galactosidase (also known as lactase, ß-D-galactoside galactohydrolase, EC 3.2.1.23).

What are galactosidase units?

There are many currently employed unit definitions for ß-galactosidase, almost all of which are designated "Miller units".

E.g., on

Promega's Web site  : the following definition is posted -  One unit of ß-galactosidase hydrolyzes one micromole of ONPG to o-nitrophenol and galactose per minute at pH 7.5 at 37°C.

On the Heber Biotec S.A. Web site One enzyme unit is defined as 1 nm of o-nitrophenol produced per minute under the following trial conditions: Substrate: ortho-nitrophenol ß-D-galacto-piranoside (ONPG) 0.2 mg/mL pH: 7.0;Temperature: 30°C

Some authors instead substitute the substrate CPRG for ONPG, and still express activity in "Miller units". Reference is available upon request. In fact, the Miller's "Miller units"were defined in nm/min, using concentration 0.66 mg/ml of ONPG, and incubating at 28°C, pH: 7.0.

What is a "normal" standard deviation in ß-galactosidase activity assays?

38% of standard deviation is acceptable for Cell; 30% makes you eligible for Nature. In our experience, 20% is as good as it usually gets in a well controlled experiment, and 10% variance reflects exceptionally good luck.
Dever, T. E., Feng, L., Wek, R. C., Cigan, A. M., Donahue, T. D., and Hinnebusch, A. G. (1992) Cell 68, 585-596
Tzamarias, D. & Struhl, K. (1994) Nature (London) 369, 758-761
 

What are the major sources of high standard deviation?

A major source is you, i.e. human inconsistency provoked by the superficial simplicity of the assay. To illustrate the overwhelming importance of this component, a simple experiment was done on two-hybrid class at NIH where one of authors (IS) was teaching. Four clones, representing negative, low, medium and high activity (lower four clones from this Figure) were grown in flasks; the OD₆₀₀ was determined by IS and posted to be used in calculations. Ten pairs of students, each including at least one Ph.D., were given aliquots of necessary reagents, printed protocols as well as detailed oral explanations of the procedure. Still, the differences in activity determined by the students varied over more than one order of magnitude (data not shown).
The second important source is clonal variation in the ß-gal expression in individual yeast colonies.  This is attributable to a number of factors, including plasmid copy number, overexpression effects, and other factors:  see our paper for more detailed explanations.
Although the recently discovered variation in the activity of ß-galactosidase molecules themselves can reach 23-fold, (see here), it could not practically contribute to high deviation.
 

How to "neutralize" the human factor?
Never split a large body of work between you and someone else. If you absolutely have to, be sure to spike the data of both researchers with several reference points. Better still, use one of the high-throughput protocols, e.g. those shown below., and STILL include several standard reference points.

Since units & conditions are so variably defined, is it possible to compare data from different sources?

In general, data cannot be compared directly. The only way to do so is to compare ratios against a known reference point(s). This is a reasonably common, though not yet (unfortunately) 100% used practice (e.g., see in ref. below or here). Using normalization, even the most dramatic differences in activity can be brought into acceptable range. For example, in the NIH experiment described above, the ratios between the activities was determined using a low-activity sample as a reference point. While the differences in activity reached more than 10-fold, the differences in ratios did not exceed 2-fold.

B. Chambraud et al. Immunophilins, Refsum disease, and lupus nephritis: The peroxisomal enzyme phytanoyl-COA-hydroxylase is a new FKBP-associated protein. PNAS 1999 96: 2104-2109. 
 

How sensitive can ß-gal assays be?
Increasing sensitivity can be obtained for increasing expense. The general order of sensitivity is,
 
 

ONPG	< X-Gal	< CPRG	< Luminogenic substrates	<  Fluorogenic substrates
3 x 108		3 x 107	4000 molecules	900 to 1 molecule

 
Fluorogenic substrates provide the ultimate sensitivity, and enable activity determination in single yeast cells and even for single molecules of ß-galactosidase. For common assay purposes, generally the first four classes of substrate suffice.
(ILYA  - can you insert the # molecules detected, from the Table in the paper).

A Single-Cell Assay of ß-Galactosidase Activity in Saccharomyces Cerevisiae." K.D. Wittrup, J.E. Bailey. Cytometry 9, 394 (1988)
Craig, DB & Dovichi, NJ (1998) E. coli ß-galactosidase is heterogeneous with respect to the activity of individual molecules. Can. J. Chem. 76, 623-626.
 

Which techniques are best suited for qualitative or quantitative assays?

Initially, in liquid assays using ONPG were used to produce quantitative data, while X-Gal based assays were intended to assess the presence/absence of activity. Recent advances in 96-plate/replicator based techniques development have provided other options. On one hand, blue spots can be quantitated and used to compare relative activities of individual clones. On the other hand, sensitive CPRG-based assays in liquid are sometimes used to qualitatively determine presence of ß-gal positive subpopulation in pools of yeast cells.  In both cases, the newer techniques help to streamline the processes.

Serebriiskii, I., G.Toby, and Golemis, E. Streamlined yeast colorimetric reporter activity assays, using scanners and plate readers. Biotechniques, revised manuscript submitted.
Buckholz, R.G., Simmons, C.A., Stuart, J.M. and Weiner, M.P. 1999. Automation of Yeast Two-Hybrid Screening. J. Molec. Microbiol. Biotechnol. 1, 135-140.
 

Which qualitative technique is better: growing on X-Gal plates, overlay or filter lifts? Growing on X-Gal plates gives a graded response that corresponds well to increasing activator strength. However, it is far less sensitive than either overlay assays or filter lifts, and thus is not suitable when low activity is expected. Overlays and filter lifts are much more sensitive: in our hands, overlays show more dose response.  Under certain circumstances, these assays can yield contradictory results; see our paper in Biotechniques P, heavy patches of yeast were put on X-Gal plates and grown for 36 hrs; O, overlay assay on plate grown for 18 hrs and inoculated with diluted yeast suspensions; F, filter lifts from a similar plate; R, ratios of ß-gal determined in liquid (lowest activity assigned 1)

What materials are the best to use for filter lifts?

Many types of carrier were tried: Whatman paper, nitrocellulose, and nylon. Various materials support development of color spots in different time frame, of different uniformity, and, surprisingly, of different shades of blue/greenish blue. The net result is almost the same, provided reaction is stopped at the right time.

What is the best way to freeze the filter?

Freezing techniques include liquid nitrogen (-196°C) and conventional deep freezers (-70°C / -80°C). Freezing in liquid nitrogen is quick, provided a source of it is at hand. On the other hand, freezing of a large number of filters is more convenient in freezers, which allow placing of large number of filters on a flat support (trays) simultaneously. Both practices yield a comparable quality of results.

What is the best way to expose the filter to substrate?

Exposure to the substrate is, probably, the most crucial step in all procedures. Limiting amounts of substrate leads to unequal access to it and distort the overall activity, while excess of liquid typically yields diffuse spots. The two most foolproof techniques are either placing filters on Z buffer/substrate-saturated Whatman 3M paper, or on "pad" of Z-buffer/substrate/0.7% agar.

Sources of substrates

A variety of colorimetric and fluorogenic substrates very inexpensive.	Diagnostic Chemicals Limited
An even greater variety of colorimetric and fluorogenic substrates at a slightly greater price	Biosynth Enter "galactopyranoside" as a search word.
CPRG sole source, to our knowledge	Roche Molecular Biochemicals
chemiluminescent detection substrates	Tropix
A great variety of fluorogenic substrates	Molecular Probes
Fluorescein di-galactopyranoside	Beckman Coulter

Although a choice of non-blue derivatives of X-Gal is available (Magenta (see example on next page), Pink, Purple, Green etc., you should excercise caution. One substrate, for example, was advertized as red... until we bought it and tried it. Now it is marked as "colorless", but otherwise equivalent derivative of other sugars are still "red"...

Commercial ß-Galactosidase Assay Kits

96-well format colorimetric ß-Galactosidase Assay	Pierce
FACS and 96-well format fluorescent ß-Galactosidase Detection Kit	Marker Gene
chemiluminescent detection in microplates	Tropix
combined luminescent detection of ß-galactosidase and luciferase	Tropix

 

Protocols on-line

Modified CHCl3- X-Gal agarose overlay protocol	Right here
A very useful collection of protocols: using ONPG, CPRG, luminescence, -X-Gal...	Clontech's Yeast Protocols Handbook (download a PDF file)
Three procols: for filterlifts, in liquid assay, and overlay	Herskowitz Lab at University of California
Filter Assay	David Bowtell's page at Peter MacCallum Cancer Institute
For Tropix products	Tropix protocol page (download PDF file(s))
Simplified protocol for colorimetric detection in 96-well plates; protocol for quantification of X-Gal overlay plates	Serebriiskii, I., G.Toby, and Golemis, E. Streamlined yeast colorimetric reporter activity assays, using scanners and plate readers. BioTechniques 29:278-288 (August 2000). (not yet on-line)
96-well format ß-Galactosidase Detection using CPRG (and, preferably, robotics)	Courtesy Vladimir Khazakhere

Back to the Yeast Two-Hybrid  Page