Hi David,

The choice of the method to analyse bacterial comunity composition also depends on your question(s) and therefore on the number of sample you plan to analyse (?).

One main difficulty you may encounter with DGGE is to standardize the method and the gradient marker from one gel to another, so that it is difficult to compare fingerprints of more than say 10 samples. Beside you may not obtain the genera and species name of the PCR product (Operative Taxonomic Unit, OTU)

qPCR methods will quantify to the level of the genera. Briefly, the method requires to establish calibration curves with target bacterial genera (decimal dilutions of target DNA from pure culture or from cloned target DNA), control for false positive (non target genera) and control for false negative (precise melting temperature and/or sequençing of some qPCR products. Once the method is established in the lab, you may run hundreds of samples... but it becomes however very expansive with an increasing number of sample!

Quite similar to tRFLF that I use to analyse fungal communities, for bacterial communities, I do prefer using ARISA (Automated Ribosomic Intergenic Spacer Analysis), a straightforward PCR method coupled to capillary electrophoresis, that only requires one fluorescent primer to generate bacterial fingerprints. Again you dont obtain the names of the OTUs, but you can run hundreds of samples. This is a very suitable method if the idea is to compare several samples between them and then to select those of interest (according to your question) and initiate pyrosequencing.

Pascal

Dear David, you are comparing two totally different methods. qPCR is for gene quatifiation and DGGE for community profiling, i.e. diversity estimation. So with DGGE you are not able to quantify your gene of interest. Best, Jiri

One can potentially use gel quantification methods in DGGE for more sensitive results.

Hi Jay, DGGE you CAN NOT use for quantification. You have PCR bias before electrophoresis. Only presence/absence of band is reasonable and accurate.

http://users.ugent.be/~trvdewie/pdf/Labmet%20publicatie%20732%20Possemiers,Verthe,Uyttendaele,VW%20%20%20%20PCR-DGGE%20based...%20FEMS%20%20Micr.Ecol..pdf

PCR-DGGE-based quantification of stability

of the microbial community in a simulator of the human intestinal

microbial ecosystem

>>>>>>>>>

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC92965/

Application of Denaturing Gradient Gel Electrophoresis (DGGE) To Study the Diversity of Marine Picoeukaryotic Assemblages and Comparison of DGGE with Other Molecular Techniques

+++>>" The reproducibility of the fingerprints was estimated by quantifying the intensities of the same bands obtained in independent PCR and DGGE analyses, and the standard error of these estimates was less than 2% on average"

>>>>>>>>>>

PCR-DGGE-based quantification of stability

of the microbial community in a simulator of the human intestinal

microbial ecosystem

http://www.biotechniques.com/multimedia/archive/00004/BTN_A_05384ST04_O_4218a.pdf

High David, I have used SSCP and have been able to get output, but I know that all these techniques have their own drawbacks. I found SSCP more favourable

As most of the people suggested, use the DGGE. Alternative could be TRFLP or deep sequencing (roche 454 pyrosequencing/illumina). Good luck!

DGGE and terminal restriction fragment length polymorphism (t-RFLP) are considered 'open-ended' approaches, since they do not target specific bacteria and so, are useful tools to examine similarities between communities, community diversity, equitability etc. We have used t-RFLP as a quick and easy tool to identify dissimilar communities and have then gone in with deep sequencing or targeted molecular approaches to identify the species contributing to the differences.

(Fullmer et al J Dent Res, Delima et al, J Clin Micro)

I also suggest t-RFLP, if you cant afford pyrosequencing of the 16S gene, which would be the best choice in fact. DGGE is semiquantitative and has got low coverage

Hi David,

The choice of the method to analyse bacterial comunity composition also depends on your question(s) and therefore on the number of sample you plan to analyse (?).

One main difficulty you may encounter with DGGE is to standardize the method and the gradient marker from one gel to another, so that it is difficult to compare fingerprints of more than say 10 samples. Beside you may not obtain the genera and species name of the PCR product (Operative Taxonomic Unit, OTU)

qPCR methods will quantify to the level of the genera. Briefly, the method requires to establish calibration curves with target bacterial genera (decimal dilutions of target DNA from pure culture or from cloned target DNA), control for false positive (non target genera) and control for false negative (precise melting temperature and/or sequençing of some qPCR products. Once the method is established in the lab, you may run hundreds of samples... but it becomes however very expansive with an increasing number of sample!

Quite similar to tRFLF that I use to analyse fungal communities, for bacterial communities, I do prefer using ARISA (Automated Ribosomic Intergenic Spacer Analysis), a straightforward PCR method coupled to capillary electrophoresis, that only requires one fluorescent primer to generate bacterial fingerprints. Again you dont obtain the names of the OTUs, but you can run hundreds of samples. This is a very suitable method if the idea is to compare several samples between them and then to select those of interest (according to your question) and initiate pyrosequencing.

Pascal

I would suggest you reconsider 454 sequencing using bar-coded primers, realistically you will end up with more and better information, especially if you can split plates with other investigators or if you have sufficient samples to run.

Hi David, I used to work for a company that did 454 sequencing for microbial diversity (16S) it was really economical, pricing started at $150/ assay at 3,000 reads. They also offer volume discounts. You might want to check them out www.researchandtesting.com

Well if you don't need to quantitate, DGGE can suffice, it can give you a good picture of the diversity of in your samples. However, should you need to quantitate, Taqman will be of help.

Hi David,

The DGGE approach would give you a reasonable comparison of the relative proportions for each peak in the sample. The same is true for sequencing for that matter.

You can follow up on the DGGE profiling by cutting the bands out of the gel and sequencing them to get a taxonomic placement. But as it was mentioned, you have to optimize the gradient for your samples, and it is pretty low throughput.

t-RFLP solves some of the throughput issues, but you can't get taxa names (usually you'd see a clone library screening in parallel to complement the t-RFLP data).

If you need absolute quantities, you'd need to use a quantitative PCR approach (either Taqman/probe based, or sybr-green based qPCR, each with its own advantages/drawbacks), or a 16S rRNA hybridization approach to count the cells (FISH and/or card-FISH). The SILVA database has some very useful information, including protocols:

http://www.arb-silva.de/fish-probes/

The sequencing approach using barcoded primers is your best cost/benefit bet. The technology is well tested for 454 based 16S sequencing.

You can also use the Illumina platform instead. See the following for a description:

Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms.Caporaso et al. ISME J. 2012 August; 6(8): 1621–1624.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3400413/

The sequencing prices have come down quite a bit and the amount of data is huge. When you add all the costs per sample involved in each method, the barcoded primer approach usually wins by a land slide, even when the number of samples is relatively small.

Best of luck!

Hi David

To generate a metagenomic profile i.e. an inventory of the genes in a mixed microbial community you need Illumina HiSeq or similar. To generate a profile reflecting community composition - which is my understanding of your question - you can use FAMES or a 16S rRNA gene amplicon fingerprinting method. I would advise distancing yourself from these approaches though as you will most likely experience harsh reviews when you go to publish as the data is less informative than that obtained by sequencing.

Pyrotag sequencing is the best option for community profiling at the moment. As pointed out by Chris Hoffmann the 454 or MiSeq platforms are fit for this purpose. Ion Torrent is error prone. Depending on your question you may not need many sequences per sample. For example the HMP and MetaHit papers are published in top journals with just 100 sequences per sample as their goal was simply to characterise the core human microbiome. Other questions may require slightly deeper sequencing - 1000 sequences per sample has been recommended as a good balance between the number of samples and the depth of sampling (Hamady and Knight. 2009. Microbial community profiling for human microbiome projects: Tools, techniques, and challenges. Genome Research 19: 1141-1152). It is worth noting, however, that it is difficult to define an abundance threshold below which microbes can be considered unimportant and metagenomics and metatranscriptomics studies are highlighting many examples were low abundance organisms hit well above their weight in terms of apparent activity.