Philip first of all which 16s primers did you use to amplify the product that you sequenced? if it is the universal 16s rRNA gene primers the product will compose of the 16s genes of all bacteria species found in the environmental sample. If you want to know the different bacteria species, you can create a clone library, then do a RFLP of random clones of the library, group clones with the same restriction pattern together and then select one clone from each group for sequencing. Then you will get the different bacteria species in you environmental sample.

Thank you Che. Yes, I used universal bacterial primers 27F and 1492R to amplify the product. Do you think that I can glean any information from my sanger sequencing result? Does it possibly indicate that Hafnia is the most abundant species as the sequence matched after blast search? Thank you

I dont thing so Philip.

Hello there. I'm afraid Sanger sequencing directly from an environmental sample isn't a wise move, especially if you are interested in the abundance of bacteria. This would be very misleading. As previously mentioned, you will need to implement cloning techniques or next-generation sequencing if you want to gain accurate information on this. Good luck!

Did you place the amplified PCR product into a vector? And picked one colony and this was the result? E.g. used TA cloning.

Or did you straight amplify and sequence from environmental gDNA?

I am not sure I understood you correctly. Does it mean that you got just one Sanger sequence from your environmental sample? If yes, the sequence you got is probably a mix of many species you have in your sample and therefore it is difficult to identify it by blast. You should not sequence 16S gene from environmental samples directly. First, you must clone your 16S product and then Sanger sequence individual clones. Cloning of 16S product followed by Sanger sequencing of individual clones is actually the same what is happening in pyrosequencing, with the difference that pyrosequencing is a highthroughput method.

You have a mix of DNA and those were just the abundant base calls. Hafnia alvei and Rahnella aquatilis may or may not even be in your sample, or it's possible they are abundant (as you have suggested). You can either reamplify with pyrosequencing or illumina primers for next gen sequencing, or you will have to clone and pick colonies (as described above) if you want to use sanger sequencing to determine who is there.

If I have understood correctly, you extracted DNA from an environmental sample and did one 16s PCR reaction on it. You then took the PCR product from this reaction and did sanger sequencing using the same primers as the PCR. You then generated a contig of the F and R sequences and then BLAST searched the consensus sequence.

If there was a mixture of bacteria in your sample you may have got mixed 16s PCR products. In this case you'd be able to see multiple overlapping different sequences in the sanger sequencing.

If there was one species in your sample you should see very clean sequence that isn't mixed at many of the bases.

However, since any set of 16s primers has a different affinity for different sequences what you amplify may be a reflection of how preferentially the primers bind to one species over another rather than the relative abundance of each species in the sample.

If the sequence you got matches both Halfnia and Rahnella I'd say that just says that these species are not distinguishable by 16s not that they are both there. If you are doing PCR on a mixed population you have to separate out each product in the mixed PCR by cloning or something or you will just get mixed sequences.

First does your sequence appear to have peaks within peaks (does it look like it is possible that two different bases could be called in a spot. This will likely show you if there are multiple species within your template. Your assumption is good that you likely picked out the most abundant species within your sample, however, you would need to do some sort of verification to prove this assumption and a good approach in NGS if you have it available and see how many hits you get to specific species, or also a clone library, although if you have two very abundant species and the others within the sample are not very abundant then you would run the risk of not identifying them with a clone library and just having many clones of the same 16s as you already identified.

Well, my scientific answer is: No, you should do some clone libraries, deep sequencing, or whatsoever to be sure.

My daily lab experience says: If you get a good sequence from directly sequencing a PCR product with those primers from an environmental DNA extract, you have a very low diversity sample or a huge PCR bias.

If you plan on publishing something, make sure you do it right. To plan your work, I would go with your assumption. I always got a messy sequence even if my culture was only contaminated with about 10% of another bug. You can see it in the chromatogram file if you have a lower abundance second sequence below the one the computer analyzed for you. You'd have smaller peaks with multiple colors in addition to the dominant ones.

I find it a bit odd that it is 100% similar to both Hafnia alvei and Rahnella aquatilis. When I did a quick comparison now NCBI tells me that there is only 96% (1427/1481) identity between the 16S rRNA gene sequences of these two species. So my first question is, are we talking about two different sequences?

Dear Philip,

the Sanger sequencing method is not a proper tool for metagenomics in an environmental sample. Due to PCR bias and poor read number I would suggest a pyrosequencing approach. Additionally you can perform DGGE or TRFLP to make a fingerprint analysis. If you want to stick with Sanger, make a clone library.

Best regards

Ingo

You should not sequence 16S gene from environmental samples directly. First, you must clone your 16S product and then Sanger sequence individual clones. Cloning of 16S product followed by Sanger sequencing of individual clones is actually the same what is happening in pyrosequencing, with the difference that pyrosequencing is a highthroughput method.

Could you do pyrosequencing on the sample?

With PCR amplification prior to sequencing I would anticipate that there would be preferential amplification of some rRNA species vs others and that this method may not give an accurate representation of species diversity.

There may be some species present at lower levels than the predominant two that you detected but the amplification of their rRNA sequences is out competed by that of Hafnia alvei and Rahnella aquatilis and so are not detectable via sanger sequencing.

Ideally you would want to deep sequence using a next gen platform such as the HiSeq and prep sequencing libraries without amplification. Kits are available for these purposes. This would remove bias and give a better representation of what is actually present in the environment. This method requires a higher input of starting material however and so may not be suitable for your specific purposes.

Dear Sir

When you are amplify 16S from environment samples means amplified products is from all organisms present in environment. In that case You have to prepare 16S r-RNA gene clone library. So, each clone contain one 16S r-RNA gene. Means each clone represents one bacterium in environment. And if you want to run DGGE then it will give ideal how many different species are present in samples. Again in DGGE You have to cut gel and sequenced If you want to identify each species which again difficult. And Now days People do NGS, Means No library preparation. You have to extrct DNA and then Run on NGS. If you are interesting In NGS i will give detail. In NGS You will good coverage.

When you do 16S rRNA PCR form any sample which contain poly microbes it tend to amplify all of them and when you try to sequence them by sanger method will get mixed peaks in electropherogram ur but i wonder how you got 100% homology, the question is how many bp is your 16S PCR product and how many base pair you finally used for BLAST since 16S highly conserved many be you have taken the region where it homologous with two spp as you got. In my experience even if you have two microbes you will get well mixed peak. If you looking for the distribution of microbes in environmental sample two best ways is DGGE or Metagenomics. These methods will give you better picture of what you looking for.

Doing 16S PCR on heterogenous sample you will get decent representation when your PCR will have only 15 cycles. That's enough to obtain high product yield.

http://www.ncbi.nlm.nih.gov/pubmed/17343679

In case the PCR reaction consist of 25 or more cycles you will get over-representation of most abundant species. Anyway to have more coverage clone library is a must.

To ensure proper identification amplification of whole sequence is needed (1.5 kb). Another issue is to secure complete elongation - do not hurry up with fast PCR. In case of producing incomplete amplicons some of them may serve as primers for different templates and that may generate chimeric amplicons. That's why it is good practice when performing multiple identifications to do chimera check

http://www.ncbi.nlm.nih.gov/pubmed/21212162

(or related papers)

Many of the above answers are in essence correct, but do not get to the heart of the matter. What is your question? Why have you amplified this sample? Once you have determined what you need to do to address your question then you need to decide which method(s) of analysis is(are) bested suited to provide you with an answer.

Ask yourself first what is my hypothesis and how can I test it. Once you can state this then picking the correct approach is much easier. If you do not have a very clear research question/hypothesis I'm afraid you will struggle to make real progress whether you Sanger sequence a PCR product or collect many millions/billions of reads using a NGS platform or create a metagenomic library.

Dear Philip,

I agree with Joerg Deutzmann and his "scientific answer". Clone libraries or deep sequencing would be the best way to get (nealy) all the species in your sample.

But for an fast overview you can optimize the PCR with group specific primers: This technique was used by Kommedal et al., 2011. They were able to design primers optimized for diferent genera and to devide them into three groups: non-anaerobic Gram-positive rods // non-anaerobic Gram-negative bacteria // Staphylococcus spp. and further anaerobic bacteria.

For the readout of mixed species chromatograms after "Sanger-sequencing" I tested an online tool, which was designed to differentiate up to three species within one 16s RNA sequence. It worked very well: ISENTIO RipSeq, Questions & Answers, Version 2.0. Paper: Kommedal et al., 2008, 2009.

I hope this was helpful. Good luck!

Thank you all for your very informative answers. I am aiming to do a microbiome analysis of the fish gut. The reason that I used a universal 16S PCR was merely to validate my DNA extraction method in ensuring bacterial DNA could be picked up. I sequenced the band I retrieved from my PCR just to see what the sequence would look like, and whether or not the most abundant species present would be represented. The sequence retrieved was 1022nt in length, with only 8 "n" base calls. This sequence most closely matched the two species mentioned above when blasted. I therefore think that either this is sample has a very low bacterial diversity or that a large PCR bias occurred.

My DNA readings for the extracted sample (before PCR) were 71ng/ul on Nanodrop, and 4.1ng/ul on Qubit, so low levels of extracted DNA. I think that pyrosequencing/DGGE is the only way to determine the true bacterial diversity in the sample, but am a little concerned that the diversity will be extremely low due to the fact that the sequence is that clean??

Thanks again

Philip

Philip

What do you want to know? Do you want to see whether different fish have similar microbiomes? This can be done quickly using DGGE but this will only really tell you whether different samples are very different from each other.

Do you want to determine what is in the gut? To do this a NGS approach is the only really valid way to do this today. 454 will do this task very well but a deeper analysis for far less money can be achieved using the illumina MiSeq even with multiplexing your samples. Both 454 and MiSeq analyses are length limited, so pick the fragment you wish to analyse carefully (perhaps follow the advice of the Human Genome Project).

However, as I said early what is your real underlying research question/hypothesis, as it is this that should drive your choice of method rather than ease of use, newness or a blind count of how much sequence you can collect!

Kevin

Hi Kevin

I want to determine whether there are a core of gut microbes (attached and transient) that play a part in digestion and overall health of fish. As DGGE will miss many of these microbes, perhaps NGS is the way forward, and has been used in similar studies of terrestrial animals and humans.

So yes, determining what is in the gut is the important part, (my research question), and my hypothesis is that there are common gut microbes (a core) that could possibly play a part in digestion efficiency and health of the fish. Again, I think that NGS platforms represent the best way of me testing this hypothesis.

Thank you very much.

Philip

You can get a hint that there are other sequences present when you look at electropherograms. The basecaller for sanger sequencing data can be set to call N when a secondary peak reaches certain percentage of primary peak area. Can be more or less stringent. Thus you can have single sequence species basecalled, but you can get a hint of other (less abundant) sequences present when additional minor peak(s) is(are) present.

If you see other peaks present, you can opt for cloning of PCR products and Sanger sequencing of reasonable number of samples (based on the height of secondary peaks), but the NGS would be a better bet.

There are various 16S primer sets and different combinations can give you different results. You can sanger sequence different 16S amplicons and check if results match.

Again and obviously, the golden standard for determination of enviromental sample composition would be the next gen sequencing. Hovewer, there are several variables that can alter the profile (targeted 16S region, PCR biases, library prep method, !DATA ANALYSIS!, etc.), but there is a great worflow described, that was adopted in a number of studies, shows good reproducibility, and should give you comparable data:

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

Then there is the question of the source sample; if/how the environmental sample was cultured...

Thank you all for your answers here. Just looking at the actual raw sequence itself, what are the signs that it is mixed? There are quite a lot of lower case base calls, could that be representative of a multiple overlaps in the sample?

Do you have a sequencing run file, or only a FASTA sequence? You need the run file, where you can view not only the bases but also their peaks. If it is a single seuquence species, a clear peak with no background will be visible. If you see some additional small peak(s) within each peak (or in some, but not all peaks), something else is there, too. Sometimes you get even level of secondary signal accross the whole sequence read, which can be a result of residual primers from PCR reaction, giving mixed sequencing products and a mixes signal. This is what you are NOT looking for. I would suggest to look for a presence of secondary peaks in only some positions with varying intesity (but watch out for artefacts such as dye blobs; they are fat peaks spreading accross more than 1 base). Maybe it helped to show your runs to some Sanger sequencing specialist in house to help you with that.

Hi Tomas

Yes I checked the trace file, and there are peaks within peaks, so I guess there's multiple species as predicted. The control sequence (from DNA extraced from a single bacterium) is a markedly cleaner sequence. I will now proceed with using other methods to distinguish the different species I have in my samples.

Thank you.

Phil

Thank you all for you help with this. I really appreciate all of the constructive feedback.

If you are able to do a DGGE analysis I'd do that. You can see if there are different species in the environmental sample by seeking different bands in the gel. After that, you can cut these bands from the gel, purify them and sequence by Sanger again, sou you can see if they are the same species you get from the PCR products or not. (probably you will get similar results, as these techiniques are less powerful than 454 or Illumina sequencing for elucidate diversity. They are much more cheaper also ;) )

Dear Philip

I agree with the overall of what has been said here. But I do disagree about putting DGGE and NGS on the same pace. While DGGE can give you a fingerprint of the bacterial population, it will not give you the same amount of data that pyro or Illumina will give. DGGE is good for visualization purposes, and easy sample comparison. It's true that you can cut and purify the bands, but you will have to clone them since more that one species can be in the same band, and you will have to sequence at least 3-4 clones of each band. So I do encourage you to do a pyro or Illumina, they are not so costy any more, even compared with the above method and will grant you more data.

The main problem that I envisage in your work is the quantity of DNA that you have. When working with environmental samples the DNA extraction step is the most important, so I advise you to use a good commercial extraction or a good manual extraction protocol that will yield high amounts and good quality DNA to start with.

Hi Philip,

I would have expected a much higher concentration of DNA in the initial genomic extract. I would try the Mobio products for this. http://www.mobio.com/fecal-sample-type/

They are a little expensive but they get great DNA recovery. It is important to start with decent amounts of high quality genomic DNA.

Also don't forget the archaea (methanogens) they will require different set of primers

Good Luck,

Emily

Dear Emily

Thank you for your reply. Perhaps you could recommend a yield that you get with this kit, as measured by Qubit perhaps?

I use the Mobio PowerSoil kit for activated sludge samples, its sister kit PowerFecal DNA Isolation is almost identical. The DNA binding capacity is 20ug total.

Considering the extracted DNA will include the genomic DNA of the fish, its food and the gut microbiome. I would aim to get ~10ug+, and elute to 50ul. Maximise the amount but be careful not to overload the kit or inhibitory substances may make it to the PCR.

I have not got my lab book with me (at work this week not uni), but from memory even the lab school students (3rd years) were able to extract original DNA at around 10-30ng/ul (eluting to 100ul not 50ul). The measurements were made using a Qbit. Some got as high as 120ng/ul but i suspect that this DNA was contaminated with something that interfered witht he Qbit measurements.

I would see if you can get some sample kits sent to you from Mobio, they are excellent. You do not need to get the vortex adaptor, when I tried them out I just taped the tubes to the existing vortex.

Hope this helps,

Emiy

Hi Emily

Thank you for your reply. I have used Qiagen kits so far, employing a bead beating pre-treatment, and obtain gDNA at 50-300ng/ul, as measured by Qubit. When measured with Nanodrop, these concentrations are obviously significantly higher.

Thank you

Philip