Hey

First you'll need to weight the organ so you can calculate the fungal/candida burden in the organ. 

Second, remove a small piece, but representative of the organ and weight that piece, always in the best sterile conditions (for example, remove the piece and put inside a sterile tube or eppendorf). 

Third, homogenise in buffer solution, like PBS. 

You'll need to check the kind of Candida your studying cause you have the type that destroys cells and so the UFC will be outiside the cell-tissue, while others prefer to remain inside the cells... so to get accurate results you'll need to evaluate that, cause if the second condition applies, you'll need to use a method that disrupts the hist cells but not the Candida cells. 

Then plate in appropriate media, like YPD or SDA, after the recurring serial dilutions. 

Hope this helps. 

Hello Javad,

Here are some answers and explanations that address your specific questions:

1) Yes. Always measure the total weight of the organ as well as the fraction of the organ. If you don't have the weight measurements and you plated a fraction of the organ, then you have no ability to quantitate the CFU and you did a valuable practice run before you do the real experiment.

The total weight of each kidney from an animal, I'm assuming mice, will vary depending on the level of inflammation caused by each C. albicans strain. Candida cells are not evenly distributed throughout the kidneys or the spleen. If you look at these tissues, there's so much Candida at the surface of kidneys that you can see the patches by eye. Attenuated strains tend to show less presence at the kidney surface and are found more toward the inside of the kidney. So I'm not convinced it's possible to obtain a representative fraction of the whole kidney if your goal is to measure differences in fungal burden from mice infected with different strains. The same goes for spleens. Liver tissue has more background bacteria than other organs and bacteria on a fungal plate can alter the number of fungal colonies that grow so I dropped liver after my first 2 experiments. I use antibiotics to prevent bacterial growth on my fungal plates from tissues. You can use 50 micrograms/ml ampicillin and 15 micrograms/ml gentamycin and use Noble SM and Johnson AD, 2005 PMID:15701792 as the reference.

The kidney is the most reliable source of CFU data from mice infected with Candida albicans. Infection with virulent Candida will often result in the enlargement of the kidneys whereas attenuated mutants will usually not cause the level of inflammation seen with the virulent strains and will be smaller. Because of these differences in total kidney size, I always measure CFU from one total kidney, either the right kidney of all mice or the left kidney of all mice and use the other kidney for other purposes. You can determine the variability among mice within each group to see if like animals have similar CFU/gram before you compare CFU between groups of mice and come to strong conclusions. Microbial pathogenesis experiments always have to be performed more than once and preferably at least 3 times to be sure the results repeat. If you want additional details about optimizing your experimental design, I'd be happy to share what I know with you.

2) To dilute a sample in "serial" means you make one dilution first, then make the next dilution with the first diluted sample, then take the second dilution and make a third dilution and so on. Each dilution is made with identical volumes such as 0.5 ml sample added to 4.5 mL PBS is a 10-fold dilution. Parallel dilutions, in contrast require different volumes of sample to be added to different volumes of PBS. A 1:10,000 parallel dilution would either require a very small aliquot of sample or a very large volume of dilution buffer. The volume of sample used for dilutions in the middle would vary for every sample and each sample size has its own margin of error. Parallel dilutions are a terrible experimental design for most assays where you want to maximize the accuracy of the procedure to minimize the differences that are caused by margins of error. Because the volumes and errors are different for the different volumes, you could be measuring differences in the CFU "counts" that are caused by error and not by biological relevance. When you treat every sample the same, homogenizing for the same amount of time, using the same pipettman to make every transfer, the margin of technical error is the same across all samples and these errors will not change the trends or log change CFU  between animals infected with different strains. Say you measure 0.5 ml but your 1 mL pipetman is off by .01 mL. If you do not change the volume setting of your pipetman between serial dilutions (use the lock mechanism), the technical error will be the same across all samples and basically drops out. Then you have a valid determination of CFU and can conclude that differences in CFU between groups of animals are due to differences in the Candida strains without sampling and pipetting errors altering your measurements. In the parallel example, a different volume of sample will have to be diluted in each measurement and will result in a different margin of error between samples. Each pipetman has a different margin of error and that error changes by range of volume. For example, the error margin for 0.1 mL taken with a 1 ml pipetman is different if you change the volume and measure 0.9 ml with the same pipetman. So always use the same pipetman and the same volume for similar measurements. If you must use different volumes for any sample dilution, go ahead and use the optimum pipetman for that volume and use it for every equivalent measurement across the samples to make any margins of error consistent across your samples.

3) To calculate CFU, you must must know the initial weight of your sample and the volume your samples were homogenized in. Example, 0.35 g of kidney in 1 ml total volume = 0.35 g tissue/1 ml and plates from the 10,000-fold serial dilution gave you reasonable numbers to count (20-200 is reasonable, 100-150 is better). The1st dilution of original homogenate = 10x, the 2nd dilution = 100X, the 3rd dilution = 1000X, an so on. The calculation can be made a few different ways. I use this one.

The # colonies counted on a plate per volume plated is multiplied by the DF and that value is then divided by the # grams tissue per ml original volume. The mL drop out of the equation and the # col/gram tissue remains.

[# colonies/ml plated X DF] / [# grams tissue/ml original homogenate] = # colonies/gram tissue = CFU/g tissue

[200 col / 0.5 ml plated x 10000] / [0.35g tissue/1 mL original volume] = 2.86 x 10^6 CFU/gram.

For graphing the data, convert to log CFU/gram.

Be sure to plate dilutions above and below the predicted dilution you will likely count so that you see the range from high to low and pick the countable plates in the middle. You might also want to plate the dilutions in triplicate and use the average per plate. Last comment, use wide-bore pipette tips. The holes in the regular tips are too small for accurate removal of an aliquot. The organ homogenate will always have chunks of material that will clog the tips. The accuracy of the entire set of serial dilutions could be significantly altered if the very first dilution is significantly different among samples. So vortex each sample a lot, use wide-bore tips, attempt to take as representative aliquots as possible from each homogenate and do it quickly before the chunks of material settle on the bottom of the tube.

Good luck!

Thanks for your comments (Diane O Inglis, Rui Filipe Ramos Santos and Richard Hilleary)

Good luck!

Hi Javad

Diane did a nice explanation about however I use a light modification in the calculation. Diane says :

[# colonies/ml plated X DF] / [# grams tissue/ml original homogenate] = # colonies/gram tissue = CFU/g tissue

[200 col / 0.5 ml plated x 10000] / [0.35g tissue/1 mL original volume] = 2.86 x 10^6 CFU/gram.

but in that way you can miss the volume that the organ occupies. If you homogenize 0.35g in 1 mL, the final volume is not 1mL anymore. I use to correct this issue assuming that 1g of tissue (specially if is from mammals) = 1 mL and then I calculate CFU/g following:

[(# colonies/mL plated x DF) x (g of tissue + mL for homogenize)] / g of tissue

However, both methods seems to be mathematically fine.

Hi Javad,

Javier pointed out an important detail about the final volume of tissue and buffer. There are so many more details that would take up quite a bit of text and time to write out for you all the details that I think are important and they include pretty much every step in the process. We haven't even touched on preparation of the inocula. If the cells aren't grown properly and prepared carefully, the entire experiment can be effected. 

I always attempted to bring the tissue up to a final volume of buffer + tissue, otherwise the CFU/gram will be an underestimate of the actual value. That is not a big problem if the comparison between CFU in infected organs is compared organ to organ, However, should you want to know what differences there are between organs within a group, each underestimation will be different between different organisms in the same group of animals. In terms of method, I like Javier's suggestion because the tissue volume can be weighed accurately on a scale and the estimation of organ volume is based a valid, precise measurement. In addition, it's more precise to deliver an exact amount of buffer volume and together those measurements are still likely to be more accurate than estimating the amount of buffer needed to bring up a tissue into a final volume by eye to a marking on the side of a tube. In that respect, I would take Javier's suggestion.

If you want some more details, an important one is to always plate a fraction of all the dilutions above and below the expected dilution that you will know the control strain yields the best dilution for counting. It's very imporatant to plate a small amount of undiluted tissue homogenate (0.1ml to 0.2 ml is fine) as well as all the DF because an attenuated mutant will have an unknown number of logs fewer CFU than your control so you must plate to the the least diluted to be able to measure the difference in CFU if the Candida is lower in abundance in a tissue and the mutant may have been cleared significantly at your time point. You will miss that information if you do not plate all dilutions. Also, there are non-biological reasons for colonies not growing on a plate and that information will let you know if there were any measurable CFU at all in each animal. If you obtain zero colonies at all dilutions, including undiluted, that fact should raise suspicion about some other source of error in the experiment that me be technical and not biological. You may have a good estimate at the dilution factor that the control plates at and your subsequent experiments should repeat with the CFU being in a similar range if you repeat the procedure in a similar manner. There are multiple reasons why an organ from an individual mouse may not have any colonies to count at the dilution factor that is optimal and the undiluted sample tells you whether any viable organisms are present at all in a mouse and in the group of mice. If there no CFU in one mouse in a group but the numbers of CFU are consistent in the rest of the animals in the same group, that tells you that the one mouse that is significantly different from it's similarly infected group, the that mouse is probably an outlier and may not have received the same amount of cells during the infection procedure. That's an indication of variability in injection technique and that can be improved with practice. If, on the other hand, nothing at all grows on even the undiluted samples in a whole group, that result carries 2 possible reasonable explanation. One is that the mutant is so attenuated that it literally clears entirely. If so, a minimum of 3 complete and carefully performed experiments with sufficient numbers of mice will be required to absolutely confirm that complete clearance, which is unusual but not impossible, is a reproducible result from a valid experiment. Another, less gratifying reason for finding zero colonies on all plates in one group but the others, is that something happened during the procedure that impacts the viability of that strain. It could be something that occurred during the preparation of cells for each group. You must always verify that each solution you prepare for injection actually has the viable number of organisms that you believe are in that tube by plating 10-fold serial dilutions of cells from the exact same tubes that contains each inoculum you use for the injections. Verifying that your starting inoculum actually has the number of viable cells that you think are present AND things can happen to batches of plates so this also verifies that all strains are capable of growing on the batch of the plates that you use to calculate final CFU. If nothing grows in beginning, or far fewer than you expected, either there issues with calculating cell number and diluting the cells, or the strain may have something about it, an auxotrophy perhaps, on that medium. So plating dilutions of the inoculum on the same plates you test for CFU and use the same buffer that you will use for tissue grinding and for making serial-dilutions.

There are experimental errors that can occur at every step in the process of carrying out infections in mice. There must be technical and procedural controls at every step of the process where something different could happen in one group and not another otherwise you may repeat the error in each experiment and without the procedural controls, you have no way to troubleshoot what the source of error may be. The procedural controls must be pre-planned and an integrated part of your experimental design for every step in the experiment in poor to be certain that the conclusion you make about the virulence properties or lack of them in a strain are truly biological because you control each and EVERY variable between animals in a group and between groups of animals and between independent repetitions of the entire experiment that gives you a result. Here's another scenario, what if the CFU are the same in the end and you have no other controls to consider, You may come to the conclusion that there's no real difference and stop there. What if you accidently used more mutant cells than the control in your experiment and you forgot to plate a dilution of the original inoculum. You won't know if a mistake occurred when the cells were counted and resuspended and too many cells were in that sample and so the error was in the sample preparation. You could be mislead by any number of different results, up or down but and will want to control for all variables and make sure that the trends and magnitude of CFU are comparable. They don't have to be exact, but the trends better be the same in replicate experiments and in a similar range of CFU if you follow the same method each time. By plating cells from the inoculum onto the same batch of plates you use to count final CFU, if just one experiment failed and the failure had something to do with the plates, plating cells before and after they pass through an animal is also important. Sometimes bad things just happen to batches of plates and cells don't grow. You want to be sure of all factors and all potential sources of error that may impact any result that you intend to claim to be true. 

A truly major source of error is in preparation of the cells for each inoculum. A few tips on that:

Grow the cells the same way. Cells from saturated overnight cultures seem to be less virulent than the equivalent number of cells in a exponentially growing culture. So you have plan the right dilution the night before and I always make an extra set of independent cultures, at the same inoculum or up or down from that inoculum for each strain, in case something happens to one culture that is different from the other cultures or to one of your sets, you should always have a back-up ready so the experiment doesn't end before it actually starts and you can continue. Use the same type of medium each time you grow the cultures for infections.

You can estimate an approximate dilutions by OD, say 2-fold greater than your desired final concentration. But the final cells for each sample must be calculated by counting accurately using a hemacytometer,Differences in size, shape, cell wall properties, etc can effect the OD and OD is not considered accurate enough for these types of experiments that are sensitive to so many variables. The preparation of inocula always took me longer than I expected and the length of time grows with the number of samples that have to be prepared. Then you'll verify that the same number of cells in a dilution produce the same number of viable cells on plates. If there are differences in viability between strains, you may need to adjust your inocula to estimated CFU/ml and verify those actual CFU/ml by plating. In any case, you need to report the viable CFU/ml for each inoculum and not just number of visible cells/ml.

I gotta go now. There are more details about the actual injections and controls for animal housing and organ collection. I'll try to get back and add some comments. Do you have any additional specific questions about any other aspect of your method that may need a control to be designed for? If so, let us know and we'll think about how to control every variable. It's very imporatnt that every single variable that change in an experiment that could lead to a incorrect interpretation must be controlled for and there are ways of controlling for every one of them by thinking every step through and planning the controls into your standard procedure.

Diane

Hi dear Diane

Thanks for detail response; I need your helps for set my thesis.

It is very essential that preparation of inoculate for injection and in many article speak about different inoculate range and it is make me confused.

Another question is that I use mutation Candida spp, in some group I don’t get culture grow, do you think it is technical problem or this species don’t pathogen?

do you think for estimate fungal burden in tissue as a replacement for this method we do real time PCR?

Another step for graphing the data you mention that it needs we convert to log cfu/gram I don’t know exactly?

regards

Dear Javad,

you have already had the right answers. I only  add that it would be a good practice that the total number of colonies that you must count for any dilution does not exceed 50-60 per plate. Above this number is easier to make mistakes due to the difficulty of seeing well each single colony. 

Regards

Alessandra Ricelli

All your comments were absolutely right. Thanks

Is it possible to convert LOD 10 CFU/ml to Molar?