Image
image
image
image

MultiTag Sequencing Platform:

Traditionally, various microbial fingerprinting methods are used to interrogate microbial communities but these techniques suffered from a lack of resolution. Alternatively, cloning and sequencing of the bacterial 16S rRNA gene has been used to identify the major components of microbial communities. Unfortunately, the methodology is slow, costly, and tedious, especially if the clone libraries are exhaustively analyzed. The main limitation is throughput as one needs to sequence about 350 clones to statistically have a 99% chance of identifying species that comprise 1% of the community. To overcome this limitation, we developed an innovative technique, Multitag pyrosequencing (MTPS), to characterize microbial communities and correlate abundance and component changes (dysbiosis) with disease states.

A new massively parallel sequencing technology has been recently introduced by 454 Life Sciences that takes individual DNA molecules and amplifies them on beads using an oil emulsion PCR technique. Individual beads are then sequenced using a pyrosequencing methodology which produces 500 million bases of sequence data from 1,000,000 different templates in a few hours. As an extension of this technology, Multitag Pyrosequencing (MTPS) uses tagged fusion primers to interrogate pooled samples, increases the throughput up to 2 orders of magnitude. In MTPS, multiple samples are barcoded using tagged fusion primers and dozens of samples are then run at the same time. Multitag Pyrosequencing allows the unique tagging and identification of multiple patient samples in a single pyrosequencing run making the technique practical and affordable for use as an IVD.

During the Phase I research, we had proposed to determine the feasibility of monitoring dysbiosis by identifying dysbiotic bacteria using fingerprints from stool samples collected on cards and rectal swabs obtained by the patient. In contrast to colonoscopy, these are non-invasive methods of collecting samples and would make it possible to more easily monitor IBD patients in order to predict flare ups and initiate therapy. We initially validated Length Heterogeneity PCR (LH-PCR) as a survey tool to monitor the dynamics of the gut microbiota. We used Length heterogeneity PCR (LH-PCR) to characterize the microbial community in 569 mucosal, lumen, rectal swab, and stool card samples. This technology offers a rapid way of screening complex microbial communities, allowing for easy fingerprinting of microfloral changes and is highly reproducible and semi-quantitative. As detailed in the Phase I results section below, we have shown that:

  • the LH-PCR fingerprinting is inexpensive, rapid and enables the screening of several hundred samples per day;
  • microbiota is stable in health in a Healthy patient;
  • the luminal and mucosal-associated microbiota are distinctly different from each other;
  • several environmental factors like fiber, alcohol consumption and sub-clinical conditions like colon polyp and diverticuli affect microbiota composition;
  • at least a portion of patients with IBD have dysbiotic microbiota.

Cloning and sequencing a complex community is an alternative approach that fingerprinting technology that actually identifies the species in the community. The main limitation is throughput as one needs to sequence about 350 clones to statistically have a 99% chance of identifying species that comprise 1% of the community (57). To overcome this limitation, we used an innovative technique, Multitag Sequencing (MTS), to characterize microbial communities and correlate abundance and component changes (dysbiosis) with disease states. MTS is an extension of the new pyrosequencing technology developed by 454 Life Sciences which produces 500 million bases of sequence data from 1,000,000 different templates with a read length of 500 bases in a few hours. The MTS technology uses tagged fusion primers to interrogate pooled samples (see below), increases the throughput up to 2 orders of magnitude, and is applicable to any environmental sample such as ocean sediments, coral communities, and gut microbiota and can be adapted to any gene of interest, such as the ITS gene for fungal identification or metabolic genes for metabolomic analysis.

It has been suggested that the 454 Pyrosequencing technology is not as accurate as classic Fluorescent sequencing, mostly due to insertions in polynucleotide tracks, and the accuracy of single pyrosequencing reads has been estimated to be as low as 96%. In contrast, it has been reported that the use of objective criteria in microbial community analysis can actually increase the accuracy rate to 99.7%. More importantly, we have observed that variation of 16S rRNA sequence of a few percent does not substantially change the identification of taxa down to the species level. Although there are competing high throughput technologies available, such as Solexa and Illumina, these technologies only produce very short read lengths (25-35 bases) and are unsuitable for bacterial identification even when using overlapping oligos. Simply put, one cannot get accurate quantitative information from a complex community with these short reads as there are too many conserved regions in the 16S rRNA and the conserved reads can match multiple species.

We routinely use the first 2 variable regions of the 16S rRNA. We have performed in silico simulations using the Eckburg dataset of full length rRNA sequences from the human microbiome and determined that one has to have 100 bp long sequence reads or longer to reproducibly characterize a community. Recently, it has been reported that most software programs for assaying microbial communities (i.e. UniFrac) are relatively insensitive to read length as long as they are over 100 bp. We routinely filter out sequences below this 100 bp cutoff for our MTPS analysis to improve the fidelity of microbial community analysis.

We have performed Multitag Sequencing™ on 135 samples to fully define the bacterial species involved in the normal and dysbiotic microbiome. As detailed in the Phase I results section below, we have shown that:

    • Multitag Sequencing™ is a reproducible technique for studying dysbiosis;
    • Multitag Sequencing™ is more informative than LH-PCR analysis;
    • The major taxa in normal controls are similar but there is variability in minor components;
    • The major taxa in normal controls is stable over time;
    • We can differentiate normal controls from apparently healthy patience with sub-clinical disease;
    • We can identify specific bacterial taxa that are associated with several pathological states;
    • Stool cards are possible surrogate samples for mucosal biopsies.



    Contact us for more info


    image


 image
image