A 4. The reason for this is that there are thousands of different strains of E. coli found in the environment, and some produce more of the one enzyme and less of the other. So if a lot of the Red Gal enzyme and only a small amount of the X-gluc enzyme is produced by a strain, the color will be more purple (and will commonly have a pink halo surrounding it), while if less Red Gal enzyme and more X-gluc enzyme is produced the colonies will be dark blue (and will commonly have no noticeable pink halo). Therefore, any colony that looks blue or purple is E. coli because that indicates that both enzymes are produced. Teal green colonies are technically not coliforms, but are likely closely related organisms like Proteus or Salmonella or an unusual form of E. coli. The only way to determine the identity of these is to run further biochemical tests, which should be done by a qualified technician or lab.

Q 5. How can I tell if a colony is really only pink and therefore a coliform, if there is a tinge of purple or blue in it (an indication of the presence of both pink and teal green) and therefore E. coli , or if it is teal green alone without any pink and therefore not a true coliform?

A 5. First of all, you should be aware of the fact that E. coli colonies appear earlier on the plate that the other coliforms. This is because the E. coli produces both enzymes and therefore a double dose of dyes which give quicker, darker color, while the other coliforms only produce the one enzyme and the lighter pink color is produced by its activity. So E. coli is visible first and has a definite blue tone, while the other coliforms appear a bit later and look pale pinkish at first. (Teal green appear later also and look very pale green with a light green halo at first.)

Second, true teal green colonies will never have a pinkish halo around them, but only a light green halo. True E. coli will either have no noticeable halo or the halo will tend to have a definite pinkish color.

If you are still not sure of which colors are associated with the colonies that you are seeing, you can purchase Micrology’s "Test Media for the Determination of E. coli from Coliforms". There are 2 different media available which, when used together, will tell you if the colony in question is producing the X-gluc enzyme, the Red Gal enzyme, both enzymes, or neither enzyme. You will need to acquire at least one set of confirmation media (#25101).  This approach is the best to use with colonies growing in/on Easygel® Coliscan® medium.

If you are using Coliscan® Membrane Filter medium, the confirmation of E. coli is very easily done by depositing a very small drop of Kovacs Reagent on the colony to be tested. This is a test for the presence of indol, which E. coli produces and other coliforms do not generally produce. Therefore, a cherry red spot will develop around E. coli colonies within ten seconds. ( See the Micrology home page for indol to see a photo.)

Last, the use of Coliscan® Easygel® Plus or Coliscan® MF will verify the presence of absence of glucuronidase in a colony.

Q 6. Do I need an incubator and how long should I incubate the plates?

A 6. An expensive commercial incubator is not necessary to obtain excellent results. However, conditions do need to be warm enough for the microbes to grow in the medium after the sample is added and the plates have been poured and gelled. Therefore, the plates should be kept in a warm area (preferred minimum temp. 80-95 F) to allow growth. This could be in a box over a hot air heat register, or better yet in a foam box where you have placed a bottle of very hot water (when the lid is closed tightly, the heat will diffuse out of the water and warm the inside of the box to a good incubation temperature and hold it for 8-10 hrs, at which point, you can renew the hot water for further incubation as needed.) Some persons make their own incubators by installing a heat tape, or a light bulb inside a foam or cardboard box, which will hold the temperature steady. However if a light bulb is used, shield the plates so the direct light doesn’t hit them and be sure that the bulb is not touching anything that it might ignite. You should try to achieve as stable, constant temperature as possible in the area of 85-95 F for most purposes to induce the fastest growth. If you are not using a thermostatically controlled incubator, you should check the plates at intervals, beginning at 24 hr, until you see some blue or pink spots developing. Then allow the plates to incubate another 24 hr and do your counts.

We highly recommend the purchase of a small, inexpensive egg incubator which works extremely well. These can often be found at farm stores for less than $40, and our favorite is made by G.Q.F. manufacturing of Savannah, Georgia. (Phone 912-236-0651, Model # 1602N Hovabator Incubator). These come with an adjustable thermostat and thermometer and we have had one operating in our lab for more than 10 years with no problems. If you are doing ongoing testing, the price is certainly right, and having one of these units eliminates any question of variations from big temperature fluctuations during the incubation period. If you set your incubator at 35 C., you should be able to count results as early as 24 hr after placing the plates in the incubator. Do not count any colonies that may develop after 48 hr as the E. coli and coliforms that you are monitoring are very fast growing and will definitely be visible within that time period.

Q 7. Is the colony size of any importance in doing my counts?

A 7. Yes. Colony size is related to several factors.

First, you need to know that E. coli and other coliforms grow very rapidly when compared to most other bacteria, so they generally appear in the plate first. Any colonies that appear after 48 hours should be disregarded.

Second, if you are incubating at lower temperature ("room temp"), the colonies will not appear as soon. In this case, you should wait until you are seeing the first colored colonies and then give at least another 18-20 hours for further growth and development of color before counting the results.

Colonies that do not grow past a pin point size (less than 0.25 mm) within 48 hrs at 35 C. should not be counted without further testing, especially if they were late showing up. An exception to this rule is if there are so many colonies in the plate that it is covered with colonies very close together. Such a plate can even look like it doesn’t have any colonies at all and just have a pink or bluish cast throughout the whole plate.

Colonies growing down in the solid medium will be smaller than those growing on the surface and surface colonies generally are paler in color. Also, if colonies are crowded close to each other on/in the medium, they will not grow as large.

Rarely, there may be present a large colored spot growing on the surface or under the medium on the plate. This is usually the result of a colony growing at the edge of the plate and down under the medium where there may be excessive moisture where it can spread, and it should be counted as only one colony forming unit.

Q 8. What is the maximum (and minimum) number of colonies that can be counted accurately on one plate?

A 8. Standard protocol for most samples dictates that Coliscan®Easygel® dishes containing more than 300 colonies (CFU) should be identified as TNTC (too numerous to count). However, it is possible to achieve quite accurate results or estimates with up to 1000 colonies/plate by careful observation. Standard protocol for Coliscan®MF dishes dictates a maximum of 80 CFU/dish due to the fact that the MF dish is a much smaller diameter (60 mm vs. 100 mm for standard dishes). It should also be noted that less than 20 CFUs/dish for either medium is considered to be statistically questionable for accuracy of count. (Note that Coliscan® Easygel® and other Easygel® media types used as pour plates in food and water testing have a big advantage over methods such as 3M’s Petrifilm® and Hach’s Coliblue® media which are limited to a small sized surface area for colony growth.)

The best way to count colonies is to position a light source so that you can easily see the colonies while looking through the bottom of the plate, and then mark each colony with a dot from a "sharpie" fine tip felt pen as you count it. That way, you avoid double counting.

Q 9. How do I know how large a sample to use?

A 9. You have to experiment or use past experience. When using Coliscan®Easygel® with surface waters such as lakes and rivers, we suggest starting with 1 mL/plate. If you know the water to be low in bacteria, you can add up to 5 mL/plate.

Water samples that have very large numbers of target organisms ( E. coli) may need to be diluted before adding to the medium. For example, if there are 500 E. coli /mL of the water being tested, the use of a 1 mL sample would result in 500 E. coli colony forming units in one plate which is too high. Therefore, a 10X dilution of the water would be desirable and 1 mL of the 10X dilution would result in 50 colony forming units/plate, which is more countable and will generally result in better accuracy. That dilution will also reduce the numbers of background organisms which is good.

If you are testing samples with very low numbers of bacteria (for example, less than 1/mL) it would be advisable to use the Coliscan®MF medium.

Q 10. How do I do a dilution?

A 10. You need a sterile measured diluent (Micrology Laboratories has diluents available, or you can make your own). The simplest way (–but not the best if you want your study to meet official protocol standards–) is to boil some distilled or deionized water to "sterilize"it. This will kill any coliforms that may be present. Measure out 90 mL in a sterile container and cool. When cool, add 10 mL of your water sample to the 90 mL of diluent, replace the lid and shake at least 25 times. Open and use 1 mL of the diluent-sample mix in your medium. When you count the target organisms, you will need to multiply the # of colonies X 10 to calculate how many there were in 1 mL of original undiluted water sample. (If there were 50 E. coli colonies growing in the plate, there would have been 500 E. coli /mL in the original undiluted water sample.)

Q 11. How can I know if information presented from studies comparing different methods is valid and that the conclusions presented with those studies are accurate and valid?

A 11. We have received or seen data and comments from numerous sources which profess to evaluate and assess the accuracy and ease of use and interpretation of our Coliscan media. While we acknowledge that some directions and experience are invaluable for interpretation of the colony colors, we believe that the Coliscan methods are second to none in their usability and accuracy. In fact, well designed studies support this thesis. Unfortunately, many persons have tried to draw conclusions from inadequate data drawn from poorly designed or executed studies. A few of the commonly encountered pitfalls follow.

A. The first factor is the lack of replications. Any comparative study should be based upon a minimum of 3 replications. This means that at least 3 identical samples from the same test sample should be run both on the Colican medium and on each of the other media with which it is being compared. Replications are necessary to determine the absence of operator error in following proper and prescribed protocol. For example, for a comparison to be valid, there should be one large test sample from which all replications for all media are drawn, and they should all be done within a short time span of each other under the same conditions by the same operator. The sample cannot be split and part of it mailed to an outside lab while the rest is used on site. (That could result in totally different viable populations of the target organisms as well as significant differences due to different operators.)

B. A very common error is trying to validate a method with inadequate data. For example, the presence of a very small number of actual colonies for the target organism/plate will generally result in large magnitudes of apparent variation. When less than ten colonies are present, a difference of 1 or 2 colonies results in possible significant variation. That is, if the sample size is 1 mL and one plate had 6 colonies and another plate had 8 colonies, the #/100 mL would be 600 and 800 respectively. That seems like a large amount, but with such small numbers of colonies, it is probably not significantly different. And if the sample size is 1 mL and one plate had 1 colony and another 3 colonies there is a 300% variation which looks very bad when it is presented in a table or graph that doesn’t note the actual numbers of colonies originally present. This can be very misleading and result in extreme, incorrect bias.

This is why most studies designed for use in approving a method will specify the minimum and maximum numbers of target colonies/plate as part of the general protocol. For example, the number for the membrane filter method is generally stated to be 20-60 /plate.

C. Another factor which may be involved in a study is variation in the types of organisms present in the test samples. Samples which are taken from the natural environment may have very large populations of one particular organism and be lacking in other types so that a study based upon samples from one or two sites may be lacking in the variety needed to test the method well. Conversely, studies based upon samples which are spiked with known varieties or organisms may not include some types that could result in false positive or negative errors so that an apparently excellent method (from the study) may be less than excellent when natural samples are used that contain more bacterial types.

These are only several factors to consider in evaluating a method and it is obvious that care must be taken to eliminate as many variables as possible so that test results are truly significant and don’t just appear to be significant and valid due to poor design, execution and interpretation of the study in question. Statistical analysis is a useful tool, but when data are manipulated or applied incorrectly due to poor initial design, the results can be very misleading. Often, what is not addressed may be more important than what is addressed. For example, one well known method advertises "low false negatives" while failing to mention that one of their own studies which was published reported over 50% false positives for total coliforms.

D. Some persons have tried to compare the Coliscan Easygel and the Coliscan MF approach/method to other methods that claim to differentiate and quantify E. coli from Other Coliforms. Two such methods are the ColiBlue® membrane filter medium of Hach and the Petrifilm™ EC medium of 3M. Each has limitations involving such things as sample size and ease of use and interpretation. One major factor of which you should be aware is that both of these media appear to be similar to the Coliscan media as they claim that E. coli appear as blue colonies and Other Coliforms are red. However, their technology must be different from that of the Coliscan media because Micrology Labs owns the exclusive rights to the patents for the simultaneous use of 2 chromogenic enzyme specific substrates to differentiate E. coli from Other Coliforms in a single medium. So instead of using X-gluc and Red Gal (Coliscan’s patented combination), they use X-gluc and a tetrazolium salt combination in their media. This means that their counts for E. coli may be accurate due to the X-gluc, but the tetrazolium salt colors all other bacterial colonies a red color, whether they are true coliforms or some other related bacteria. This is why they get no different shades of blue/purple with their E. coli and why all their "coliforms" are a uniformly dark red color, which looks very nice on the plate. They claim to include inhibitors which prevent non-coliforms from growing, but this may not be entirely possible as there are numerous bacterial types which are non-coliforms that will grow in virtually any environment suitable for coliforms. As a result, these "copycat" media may be subject to high false positive errors in the identification of total coliform counts and general coliforms. (In a 1997 publication authored by Grant of the Hach Company, he admitted to 52% false positive results for total coliforms with the m-ColiBlue 24® medium.)

Studies need to be carefully examined to determine that they lack design and data flaws, and when a method is tested, technicians should be sure they understand the technology and any potentially negative flaws. For example, microbes may be inhibited by tetrazolium salts in a medium. Work done at Micrology Labs comparing identical media and conditions, but adding tetrazolium salt to one, have resulted in generally lower recovery of organisms on the medium containing the tetrazolium salt.

Q 12. There was a notice on the box of Coliscan that said it must be frozen for storage, but when I opened the box, the medium was thawed and liquid. Do I have a problem?

A 12. No, you do not have a problem. The medium can be thawed for 2-3 weeks at room temperature with no adverse effects. Just put the bottles of liquid Coliscan in the freezer and they will be fine. Also, if you happen to thaw too many to use, they can be re-frozen without any problem. We have kept Coliscan for a month at room temperatures without noticeable negative effects, but for long term storage life, it is best frozen as the chromogenic substrates slowly lose their brilliant color over time at high temperatures.

Q 13. The pretreated plates which I received with my bottles of Easygel® look strange. Some have water droplets covering the surface of the film on the inside bottom of the plate which gives the layer a hazy appearance. Some may also have water droplets collected in the lid of the plates in some sleeves. Is this contamination or are these plates OK to use? (Conversely, you could receive plates which are so dry that the layer on the inside bottom has the appearance of frosted glass.)

A 13. No, this is not contamination, but a result of shipping and storage conditions which caused a big temperature swing within the sleeves of plates, resulting in condensate forming from the water in the pretreatment layers. When this does happen, it is usually minimal and all the moisture is retained within the individual plates involved, so there is no introduction of contaminants into the plates. The plates can be used with confidence that they are not contaminated . On rare occasions, conditions get so extreme that excessive condensate runs out of the plates and causes the inside of the plastic sleeve containing the plates to become wet. Even though this environment should still be sterile, it only takes a single spore to grow and multiply so that the entire sleeve can become contaminated (worst scenario), and we recommend that you notify Micrology Labs and request replacement if this wetting of the entire inside of the plastic sleeve occurs. . (The dry, frosted appearing plates will perform well and your results will not be adversely affected. This happens when they are stored in very dry conditions.)

Q 14. Coliscan® MF is approved by the EPA as a parallel method the the MI Method for testing drinking water. How do these methods compare?

A 14. Coliscan® MF is designed to test for the presence of E. coli and General Coliforms in liquid samples which are able to pass through a membrane filter (0.45 micrometer pore size). The method is particularly excellent for checking large samples which have low populations of these bacteria (such as drinking water or other high quality waters or other materials). It is also excellent for testing samples which have higher bacterial populations, but the sample will require dilution in such situations so that the number of bacteria on one filter does not exceed countable limits.

Coliscan® MF tests for the identical entities (galactosidase and glucuronidase) that are used as indicators in the MI medium described in papers published from work done at the National Exposure Research Laboratory of the US Environmental Protection Agency in Cincinnati, Ohio. As a result of those studies, the method reported was stated to be acceptable and recommended for water testing. "Since the results of this study show MI agar to be equal to or better than the current method, it meets the criteria for routine compliance monitoring of drinking water"(3). The MI method and the Coliscan® MF method are the same as far as mechanism and products tested for, but differ due to the specific substrates used in testing for the specific enzymes. As shown in the following explainations, the Coliscan MF method has specific advantages over the MI method.

The MI method uses small plates containing 5 mL of the agar formulation. This approach is preferred instead of the alternative of using a liquid formulation without agar which can be added to an absorbent pad in the plate because the MUGal ( indicator for the presence of the enzyme galactosidase) is soluble and would spread over the whole plate if one colony tested positive. This is still a problem with MI agar since the MUGal will still diffuse out of colonies that are galactosidase positive, although at a slower rate than if a liquid medium were used. This can cause a problem on the MI agar plates if there are many galactosidase positive colonies as well as some non-galactosidase positive colonies since the fluorescence from the MUGal may diffuse through the non-galactosidase positive colonies and cause them to appear positive.

This is not a problem with the Coliscan® MF method. It can be used either with an agar formula which is poured into the plates or as a liquid formulation which is added to an absorbent pad in the plate. The reason that there is no problem with the Coliscan® MF is that instead of using MUGal for the galactosidase determination, a water insoluble galactosidase substrate containing a pink/red chromogen is used. This results in the galactosidase positive (coliform) colonies growing with a pink/red color that does not diffuse throughout the medium.

Another advantage of the Coliscan® MF method is that there is no need to view the bacterial colonies under long wave UV light. The colonies are easily seen as blue or red colored dots in normal room lighting or with the aid of a dissecting microscope. The confusion of interpreting the presence or absence of fluorescence in specific colonies due to the diffusion of the fluorescent product from E. coli colonies to adjacent other coliforms is avoided. Therefore precise quantification is easier and more achievable. (It should be noted that the new Coliscan® MF Plus contains MUGluc for the purpose of giving a double verification of the P/A of E. coli.)