CLC bio

Identifying genetic variance related to human diabetes mellitus

CLC bio — Mon, 26 Mar 2007 20:54:00 GMT

Diabetes mellitus is causing major health problems in the world and is affecting more and more people. Today nearly 200 million people have been diagnosed, and the number of new incidences is rapidly increasing, especially in the western world. The reason for this increase is not simple. Change in lifestyle and environmental factors are important factors as indicated by many clinical studies.

In general, diabetes mellitus is a severe life threatening disease. The secondary complications are many and well described. Examples are highly increased risk of obesity, development of retinopathy, development of kidney dysfunction, and accelerated development of atherosclerosis and mucormycosis.

Diabetes mellitus types 1 and 2 are characterized by primary hyperglycemia due to dysregulation of the plasma glucose titer. Glucose uptake by somatic cells is regulated by insulin. Insulin is produced by the ß-cells in pancreas and dysfunction or apoptosis of these cells are affecting insulin expression and subsequently the plasma concentration of glucose.

Studying diabetes mellitus

The understanding of the genetic basis of diabetes mellitus and the influence of genetic variation on phenotypic traits are of great importance for medical research and development of drugs. The use of sequence analysis tools like the CLC bio workbenches are of great importance today to gain a better understanding of the complexity and of the diabetes pathogenesis.

In pre-clinical studies many different animal models have been used. The pig is a valuable model organism and is relevant to human biomedical research in areas as e.g. obesity, cardiovascular disease, and nutritional studies.

This case study describes how the CLC Combined Workbench is used to identify sequence polymorphisms in the pig genome at positions that may be related to diabetes and subsequently demonstrate how to analyze those using bioinformatics tools.

Genetics of Diabetes Mellitus

Diabetes mellitus is a complex, multifactorial and polygenic disease, likely to be caused by one or more gene alterations acting in combination with non-genetic factors [Morwessel, 1998]. Since obese phenotypic traits often are seen in relation to a diabetic diagnose, genetic analysis of genes (including sequence analysis) known to be related to obesity could be interesting for the clarification of some phenotypic relations to the disease.

Identifying uncoupling proteins as diabetes type 2 candidate genes

The candidate gene approach for type 2 diabetes mellitus tests for association between particular gene variants and diabetes. So candidate genes encode proteins involved in either insulin synthesis, pathways of insulin secretion, or insulin action, where defects cause abnormal patterns [So et al., 2000] and polymorphisms in these genes may be important risk factors for type 2 diabetes mellitus patients.

Uncoupling proteins

Uncoupling proteins (UCP) are located in the inner mitochondrial membrane, and one of the suggested functions of UCPs is that of uncoupling by acting as a channel for proton entry into the mitochondrial matrix.

Figure 1: Location of the uncoupling protein in the intramembrane space [Gura, 1998].

Figure 2: The yellow arrow represents the uncoupling protein's function as a channel for proton entry into the matrix [Rousset et al., 2004].

Respiration and ADP phosphorylation in mitochondria are coupled, and the uncoupling proteins appear to be controlling the level of these functions. Uncoupling is when a protein acts as a proton carrier, and by the transportation of protons from the intermembrane space to the matrix a shunt between ATP synthase and the respiratory chain is created [Rousset et al., 2004]. By this mechanism UCPs might have some basic roles to play in human physiology. In addition UCP2 and UCP3 decrease membrane potential and increase thermogenesis [Dalgaard and Pedersen, 2001] and the genes encoding these proteins are thus regarded as candidate genes for studies of the diabetes type 2 often accompanying obese phenotypic traits.

As the uncoupling proteins are highly conserved among species and significant similarity is seen between e.g. the human and the pig uncoupling proteins, showing more than 90 percent identity at the level of amino acid sequence, these genes might be interesting related to the animal model research.

Work flow

CLC bio provides bioinformatics software of great importance for the detection, identification, and characterization of polymorphisms in diabetes type 2 candidate genes as e.g. the uncoupling proteins. The use of CLC Combined Workbench integrating all analyses in one program is exemplified below.

Human and porcine DNA encoding UCP2 and UCP3 is retrieved from local and on-line databases using BLAST. The sequences are aligned to check for proper identity between species. Next, primers are designed for porcine DNA. The primers are used for PCR, and the products are subsequently sequenced. The sequencing data is assembled to the reference sequences used for designing primers, and putative polymorphisms are identified. Using the integrated SNP annotation functionality of the Combined Workbench, the possible polymorphisms are characterized and compared to known SNPs in the SNP database. Next, the coding regions of the DNA sequences are translated into protein and subjected to a number of predictions to determine the impact of the polymorphisms on the UCP proteins.

In the next sections, three of the steps in the work flow are described in further detail.

Zooming in on Annotations

During the entire work flow, the same sequences are used for both alignments, as basis for primer design, and as reference sequences in assembly and SNP identification. This means that annotation of genes, coding regions etc. are preserved during all the analyses. They can then be used to guide the inspection of alignments and BLAST hits, the location of primers, the interpretation of sequencing data etc. Small snapshots of the role of annotations in the different parts of the work flow are shown below:

Figure 3: Inspecting the result of a BLAST search. The yellow annotation represents the coding region of the UCP3 gene. The annotations make it easy to get an overview of where the hits align to the query sequence.

Figure 4: An alignment with a yellow annotation representing the coding region of the UCP3 gene for humans(top) and pigs(bottom). The translation is shown to visualize differences in the amino acid sequence between the two species.

Figure 5: Designing primers to bind just before the coding region of the UCP3 coding region. The annotations eliminate the need for remembering positions.

Figure 6: Assembling to the reference sequence where both the primer binding site (red annotation) and the coding sequence (yellow annotation) are shown.

Zooming in on SNP annotation using BLAST

Sequencing data of genes encoding the uncoupling proteins are searched for polymorphisms. In positions where a polymorphism is identified, a BLAST database search is performed for SNPs in human genes similar to the genes of interest. This helps control and verify the results.

Results of the SNP annotation using BLAST can be shown as graphics (see figure 8), in a tabular view, and as annotations on the input sequence(s).

Figure 7: The graphical view of the SNP BLAST. At the top you see the sequence used as reference in the assembly with possible SNP's annotated with red arrows. Below are the hits from the SNP database. One SNP is high-lighted (M).

When annotating SNPs using BLAST, you can select which database you want to search against, e.g. human, mouse, or rat. You can also set the BLAST parameters such as filtering and gap costs.

In the graphic view you can see sequence matches between your query sequence and hit sequences in the database chosen. From here you can easily zoom in on specific regions of interest, or you can open a hit region in a new view.

Annotations on the query sequence will indicate any match from the database where a polymorphism has previously been identified, and the tabular view provides an overview of e.g. identity and positions of matching regions between query and hit sequences. From the tabular view you can easily open any of the hit sequences at the NCBI web page.

Zooming in on transmembrane helix prediction and secondary structure prediction

After translation of the sequenced genes and adding of annotations where polymorphisms were identified, transmembrane regions in the porcine uncoupling proteins 2 and 3 are predicted by CLC Combined Workbench to localize the identified polymorphisms; transmembrane location or in extracellular or intermembrane regions.

In a similar way, secondary structure of the proteins were predicted in CLC Combined Workbench. From suggested locations of the polymorphisms the complications with specific domains affecting protein structure and function are identified.

As related to uncoupling protein 3 putative topology in the inner mitochondrial membrane, a suggested location of identified polymorphisms may be helpful to predict possible impacts of the identified genetic variance.

Figure 8: Prediction of secondary structure for the analyzed uncoupling protein 3.

High-performance bioinformatics without any hassles

CLC bio — Wed, 14 Mar 2007 19:27:00 GMT

CLC bio announced the release of an innovative product to the scientific community - the CLC Bioinformatics Cell - which will enable users to transform their ordinary Intel- or AMD-based workstation computers to super-computing units. Tasks previously taking two hours can now be computed in one minute - without any hassles.

Jörgen Forsberg, Nordic Director at Intel:

This product can potentially help pharma and biotech corporations move faster through the phases of drug discovery, and thereby help the pharma industry to develop products faster to meet the healthcare challenges of today

Senior Scientific Officer at CLC bio, Dr. Jannick Bendtsen adds:

A very important research area of molecular biology today concerns the use of small interfering RNA (siRNA). The design of siRNA relies on database matching to ensure that strong cross-hybridization to non-target genes does not occur. However, due to the small size of siRNA molecules, potential cross-matches can be missed by heuristic algorithms like BLAST. With the speed of the Bioinformatics Cell we are now making it possible to abandon heuristics, and to weed out cross-matching siRNAs, by using Smith-Waterman - the most accurate matching algorithm that exists.

It is commonly accepted throughout the scientific community, that the growth in amounts of data which needs to be computed is growing faster, than the increase in processor speed. This means scientists are craving for alternative ways to have their computing needs met. The CLC Bioinformatics Cell provides this, by accelerating bioinformatics algorithms, such as Smith-Waterman and ClustalW, up to 110 times faster than what a fast desktop computer can handle today.

The Bioinformatics Cell has been developed by CLC bio's world class bioinformatics experts, and they have essentially brought high-performance computing to everybody with this product. This remarkable speed-up is achieved through innovative use of SIMD technology that unleashes already existing power in your computer processor.

A scientist can buy a single unit for his laptop computer - a lab can share a couple among their desktop computers - or - it can be plugged in to each computer in a computer cluster and see the speed-up soar beyond anything available in the market today.

The CLC Bioinformatics Cell works with a simple hassle-free Plug 'n Play concept through CLC bio's intuitive workbench platform, thus eliminating the need for highly specialized consultants. The device is simply plugged in to the USB port of any Intel- or AMD-based computer, be it Mac OS X, Windows or Linux. Command-line execution is also an option, giving the option of integrating the Bioinformatics Cell with existing bioinformatics workflow of scripts and in-house programs.

Version 0.91 of the Bioinformatics Cell is currently supporting the bioinformatics algorithms of Smith-Waterman and ClustalW, with more to come. The acceleration of Smith-Waterman is such, that it can be compared to BLAST, however yielding 100% precise results instead of the imprecise estimates of BLAST, where up to 50% of the results can be missed, as stated in the white paper "High performance computing, using CLC Bioinformatics Cell", available on www.clccell.com. The CLC Bioinformatics Cell can help researchers and scientists worldwide get better scientific results in less time

Bioinformatics Consulting

CLC bio — Wed, 14 Mar 2007 19:11:00 GMT

More and more public research organizations, biotech companies, and pharmaceutical corporations need assistance in achieving specialized, faster, and more effective research results. Our world class bioinformatics consultants can help you achieve these goals and get even better scientific results - in less time! We can help you succeed by:

Developing specialized bioinformatics algorithms
Creating customized modules for your bioinformatics software
Performing bioinformatics analyses for organizations that lack time or qualified knowhow.

CLC bio Consulting - World Class Bioinformatics Consulting

Screening the carrot genome for MADS-box genes

CLC bio — Wed, 14 Mar 2007 18:57:00 GMT

This case study shows how bioinformatics in general, and CLC Combined Workbench in particular, can help researchers gain deeper knowledge of the complex processes involved in flower development in Daucus carota.

Figure 1: Flower of Daucus carota.

The study focuses on how the alignment-based primer design functionalities of the CLC Combined Workbench make it possible to design degenerate primers to amplify a set of conserved genes in Daucus carota. The research has been carried out at the Laboratory of Gene Expression at the University of Aarhus, Denmark.

Early flowering in Arabidopsis thaliana

Flowering plants, or angiosperms, represent one of the most successful and diverse groups of organisms on the planet, with more than 250,000 species in the wild. Although angiosperms such as orchids, roses and snapdragons have very distinctive flowers, most flowers contain just four organ types and their development involves highly conserved molecular mechanisms [Krizek and Fletcher, 2005].

Flowers, like shoot and leaves, derive from the shoot apical meristem (SAM). Whereas leaves arise early in the plant life cycle, flowers are only produced after transition of the SAM from vegetative to reproductive growth. The timing of the transition is crucial to reproductive success and as such must take place at the exact right time.

In Arabidopsis thaliana the switch from vegetative to reproductive growth is governed by four main pathways (see figure 2). Signals transmitted by genes belonging to each pathway converge at the Floral Pathway Integrators. These activate and/or up-regulate the expression of the Floral Meristem Identity (FMI) genes responsible for the establishment of the floral meristem [Simpson et al., 1999]. The FMI genes in turn activate the Floral Organ Identity genes, required for specifying the identity of the floral organs [Krizek and Fletcher, 2005].

Figure 2: Flower Inducing Pathways. The Gibberellic, Autonomous, Vernalization and Photoperiodic pathways induce the transition of the SAM from vegetative to reproductive growth and hence early flower development. Genes written in boldface belong to the MADS-box gene family.

Of the 15 genes central to early flower development, nine belong to a group of homeotic genes known as MADS-box genes, sharing a conserved region of approximately 180 bp in the 5'-end. Due to the high degree of conservation, degenerate primers targeting this region can be designed. This enables a broad search for MADS-box genes in virtually any organism.

Expressed in vegetative as well as floral tissues, MADS-box genes are likely to be involved in many aspects of plant development. They have been successfully cloned from more than 40 different plant genera [Purugganan et al., 1995]. In the Arabidopsis thalianagenome alone, more than 100 sequences have been recovered and although less than 20% are characterized, many have been shown to be involved in several aspects of bolting and flowering [Parenicová et al., 2003], [Martinez-Castilla and Alvarez-Buylla, 2003].

MADS-box genes in carrot (Daucus carota)

Cytoplasmic male sterility (CMS) results from incompatibility between the nucleus and cytoplasm. CMS plants are not capable of normal pollen development, in that maturation of viable pollen is prevented. In addition, flower phenotype is altered. The phenotypes resemble homeotic flower mutants of Arabidopsis thaliana.

In carrot, five nuclear genes involved in the expression of two CMS phenotypes have been identified. The five clones were aligned with putative equivalents of other plant species, and due to their homology to homeotic genes of Antirrhinum majus and Arabidopsis thaliana the genes were named DcMADS1-5 [Linke et al., 2003].

The research performed in the Laboratory of Gene Expression at the University of Aarhus alternated between tasks performed in the lab and bioinformatics tasks using the CLC Combined Workbench. Below, the work performed is summarized.

Research

The purpose of the research was to identify genes involved in the floral initiation process in carrots (Daucus carota cv. 'Nantes') and focus was set on the MADS-box genes. The work flow is outlined in figure 3.

Figure 3: Illustration of work flow.

The CLC Combined Workbench was used to search for, import and align the cDNA sequences of DcMADS1-5. Then, a pair of degenerate primers were designed targeting the conserved MADS-box region. These primers were used for PCR amplification of genomic DNA isolated from leaves of 19 weeks old carrots. The resulting fragments were cloned into the TOPO vector and sequenced.

The trace files were imported into the CLC Combined Workbench, where the forward and reverse reads were assembled. BLAST searches against the NCBI database were performed using BLASTx which translates the sequence into amino acid sequences before BLASTing against a protein database. Based on the results of the BLAST searches, the clones suggested to be fragments of MADS-box genes were selected. Based on the nucleotide sequence, primers for inverse PCR (iPCR) were designed.

Back in the lab, the positive clones were labeled with ³²P-dATP and used to screen a carrot genomic library. Positive λ-clones were isolated, DNA extracted and subjected to iPCR. The resulting amplicons were cloned into the TOPO vector and processed as described above.

The screening of the genomic library remains a work in progress and at this point in time putative MADS-box genes have been selected and annotations describing the regions identified in each clone were manually added. The images generated within the CLC Combined Workbench have been exported in high-quality graphics formats for use in presentations etc.

In the following sections, two of the steps in the work flow are described in further detail; The alignment-based primer design and export of images.

Zooming in on Alignment-based primer design

The CLC Combined Workbench makes it possible to align several sequences and design degenerate primer sets in regions of homology.

Figure 4: Alignment of four sequences which are used to design primers. The red color indicates fully conserved residues and the blue color shows positions with high degree of variation. At the positions with no background color, one of the sequences is different from the rest.

Annotations from the original nucleotide sequences or alignment are still visible within the primer design viewer, which is a big help when the primers are required to lie within specific regions. In figure 4, you can see the yellow annotations indicating the coding regions of the MADS-box genes.

The conservation is shown as a blue-to-red gradient background color making it easy to visually find the highly conserved regions of the alignment. These regions are defined as locations for the forward and the reverse primers as shown with the gray annotation in figure 4.

The parameters regarding the individual properties of the primers are adjusted in the Side Panel. In addition to normal primer settings, such as primer length, melting temperature, buffer properties and degree of self annealing, the alignment-based primer design makes it possible to select degree of degeneracy, or the number of mismatches acceptable.

When all the parameters are set, the CLC Combined Workbench calculates the best combinations of primer pairs for the selected regions. The properties for the primer pairs, such as melting temperature and self annealing, are inspected before deciding on a primer pair.

Zooming in on Export of images

In figure 5 you can see the graphical result of the first phase of the research described in this case study.

Figure 5: Graphical output of the CLC Combined Workbench. The result of the research is shown as annotations.

The graphics generated in the CLC Combined Workbench are easily exported in a variety of graphics formats such as gif, jpeg, png and pdf.

The CLC Combined Workbench graphics export is based on a WYSIWYG principle: What You See Is What You Get. This means that you can configure the layout of e.g. a sequence or alignment in the program and when you export this to a graphics file, it will look exactly the same as in the Workbench. This makes it easy to create good-looking illustrations of research findings for publications and presentations.

The annotations on the sequence shown in figure 5 illustrate putative MADS-box genes and the annotations furthermore describe the regions identified in each clone. This illustration is used for communicating the results of the research in reports and presentations.

CLC bio joins Microsoft in the BioIT Alliance

CLC bio — Thu, 25 Jan 2007 21:58:00 GMT

Aarhus, Denmark -- January 25, 2007 -- CLC bio today announced that it has joined the BioIT Alliance, a cross-industry group which unites the pharmaceutical, biotech, hardware and software industries, to explore new ways to share complex biomedical data and collaborate among multi-disciplinary teams to speed up the pace of discovery in the life sciences.

CLC bio CEO Thomas Knudsen is excited about the collaboration:

It is an honour for our company to join a network of such highly estimated companies as Microsoft, Hewlett-Packard, Sun Microsystems and others. We think this agreement will prove beneficial for all involved parties, and we’re looking forward to working together within this scientific community, developing innovative solutions and solving current issues in the post-genomic age.

Microsoft Platform Strategy Advisor Don Rule adds:

We are thrilled to welcome CLC bio to the BioIT Alliance. They bring deep experience in bioinformations software, hardware and consulting solutions to our rapidly expanding community.

About the BioIT Alliance
Formed in 2006, the BioIT Alliance is a cross-industry group working together to improve biomedical information technology on the Microsoft platform. Founding members include Affymetrix, Inc., Accelrys Software Inc., Amylin Pharmaceuticals, Inc., Applied Biosystems, The BioTeam Inc., Digipede Technologies LLC, Discovery Biosciences Corporation., Geospiza Inc., Hewlett-Packard Development Company, InterKnowlogy, Microsoft Corporation, Sun Microsystems Inc., VizX Labs LLC and other key companies in the pharmaceutical, biotech, hardware and software industries. Additional information about the BioIT Alliance can be found on the BioIT Alliance Web site at http://www.bioitalliance.org

About CLC bio
CLC bio is the world's leading full-service bioinformatics solution provider, solely focusing on the development of bioinformatics: software, hardware, data analysis, and custom-designed bioinformatics algorithms.

CLC bio’s mission is to be among the most innovative bioinformatics companies in the 21st century. This is realized through:

Development of bioinformatics software and hardware based on the latest scientific findings
User-friendly, integrated and intuitive software solutions
Continuous focus on customer needs and superior customer service
Frequent product updates including the latest IT technologies and bioinformatics algorithms
A flexible IT architecture, enabling customers to buy or develop individualized solutions at a reasonable price.

# # #

For further information, please contact:
     Thomas Knudsen, CEO
     CLC bio
     Gustav Wieds Vej 10
     8000 Aarhus C
     Denmark
     Phone: +45 70 22 55 09
     E-mail: info@clcbio.com
     Website: www.clcbio.com

Micro sites

CLC bio — Mon, 15 Jan 2007 19:06:00 GMT

CLC Bioinformatics Cube - hardware-acceleration on your desktop

CLC Developer Kit - software developer kit (SDK) for bioinformatics software plug-in developers

CLC Bio Consulting - World Class Bioinformatics Consulting

Online workbench presentations

CLC bio — Mon, 15 Jan 2007 19:05:00 GMT

Now, you can learn more about your workbench of choice while sitting comfortably at your computer in your office.

The one hour presentation is performed by one of our product specialists and transmitted to you through the internet. During the whole session, you can ask questions through a free/low-cost teleconference connection, ensuring that you do not leave with questions unanswered.

It’s very easy and without any obligations: You do not have to prepare, download or install anything on your computer – just sign up, and we will make sure that you will get a lot of valuable information.

Follow this link to read more

Alignment speed and quality at CLC bio

CLC bio — Mon, 15 Jan 2007 18:53:00 GMT

The alignment algorithms used in the software from CLC bio A/S has some unique features including the option of adjusting the cost of gaps in the end of the alignment to suit the sequences being aligned.

We have two alignments: A standard algorithm that is 10 times faster than our previous
alignment in most scenarios, and an additional alignment that is even faster, but less accurate than
the standard algorithm.

The White Paper below forms the basis for these 5 conclusions:

On large data sets of sequences that are not too divergent, our alignment is significantly faster than the standard CLUSTAL W alignment, and around the same speed as the fast CLUSTAL W alignment.
Performing an alignment of 28 HIV genomes, our fast alignment is more than 10 times (55 minutes) faster than the standard CLUSTAL W alignment.
We have benchmarked our new algorithms on the BAliBASE 3.0 database of accurate protein alignments (Thompson et al., 2005). This shows that our alignment algorithm is about 1% more accurate than the latest version of the standard CLUSTAL W on protein alignments.
We have bechmarked our new algorithms on the BRaliBase II database of structurally aligned RNA. Here, our new algorithm is about 3.5% more accurate than the standard CLUSTAL W.
Our standard algorithm is still a little slower than the standard CLUSTAL W on the fairly divergent alignments in BAliBASE and BRaliBase. Our fast alignment is as precise and as fast as the standard CLUSTAL W on these data sets.

Download White Paper

NASA Chooses Sequencing Software from CLC bio

CLC bio — Mon, 15 Jan 2007 18:46:00 GMT

The Exobiology Branch of the NASA Ames Research Center in Moffett Field, California, has chosen bioinformatics software from CLC bio for analyzing sequencing data which revolves around hypersaline microbial mat systems that are found worldwide in exotic locations such as Baja, Mexico, and Solar Lake, Egypt, to name a few.

Exobiology - also known as astrobiology - is an interdisciplinary field, combining aspects of astronomy, biology and geology, which is focused primarily on the study of the origin, distribution and evolution of life. Microbial mat communities represent, in gross morphology, some of the earliest known microbial communities on Earth. As a common form of microbial community existing on Earth during the Precambrian, and dominant in the Proterozoic era, these systems are hypothesized to have played a significant role in the development of modern oceanic and atmospheric conditions.

Dr. Stefan J. Green, scientist from the Exobiology Branch of the NASA Ames Research Center, states:

We are currently employing various molecular tools to characterize shifts in community structure of the total microbial population as a result of alterations in sulfate and salinity levels. Our initial molecular analyses are performed using a fairly standard DNA extraction-PCR amplification-denaturing gradient gel electrophoresis (DGGE) methodology. We are employing a wide variety of primers for PCR-DGGE analyses and while many of these primer sets amplify regions of the 16 or 18S rRNA gene, we are also exploring a variety of primer sets to target functional genes. Such functional gene analyses can provide information regarding microbial function and phylogeny. CLC Combined Workbench is very useful in our manipulations of sequence data recovered from our exobiological analyses. In particular, the software has been useful in assembling complete and long gene sequences from multiple sequence reactions - with the sequencing trace visible for all overlapping reactions. For example, the dsrAB genes are approximately 1900 bp in length, and require a minimum of 3 sequencing reactions. The gene annotation feature is also highly beneficial, by helping establish, for example, where the dsrA gene ends and the dsrB gene begins, based on other sequences in Genbank. The wide variety of other tools in CLC's software are extremely useful – BLAST searches from within the program, in silico restriction digests, alignments, rudimentary phylogenetic trees, among a lot of other features. One of the worst facets of sequence analysis is the constant reformatting that must be performed to analyze the data with different programs - and with CLC's software the need for many of these manipulations are eliminated. Thank you for creating such nice software!

Learn more about exobiology

Click here to read the full press release.

Detection of genomic changes in Mycobacterium tuberculosis

CLC bio — Mon, 15 Jan 2007 18:10:00 GMT

Detection of genomic changes in Mycobacterium tuberculosis

Resistance development in Mycobacterium tuberculosis

Tuberculosis (TB) remains the leading cause of death due to bacterial infections worldwide. About 8 million new cases of active TB arise each year resulting in 3 million annual deaths. Roughly 1 billion individuals are believed to harbour latent tuberculosis.

Primary infections with M. tuberculosis are generally asymptomatic but will in some cases remain as a latent infection.

Tuberculosis is a secondary disease caused by reactivation of the M. tuberculosis bacteria not fully eliminated after the primary infection.

Sequencing and sequence analysis of bacterial and fungal genomes and proteins/peptides have led to a better general understanding of the pathogenesis of bacterial and fungal infections. Future understanding of the regulatory events at the molecular level will increase and be accelerated by using a variety of new technologies and technology platforms within microarrays, protein chips and sequence analysis tools (CLC bio workbenches). The aim is to develop more specific and effective drugs much faster to target e.g. the expanding multi resistance Mycobacterium tuberculosis strains in human populations all over the world.

Multidrug resistant bacterial strains arise by sequential accumulation of resistance mutations for individual drugs. A diverse array of strategies is available to assist in rapid detection of drug resistance-associated gene mutations.

Bioinformatics (CLC bio workbenches) together with functional genomics and functional proteomics have been used to identify expression pattern (signature) changes in multi resistant M. tuberculosis strains.

Mycobacterium tuberculosis genome

The Mycobacterium tuberculosis genome was sequenced in 1998 [Cole et al., 1998] (see figure 1). Today many strains of M. tuberculosis have been sequenced. In pest control laboratories, sequencing is routinely performed to monitor strain genomic changes.

Figure 1: The Mycobacterium tuberculosis genome [Cole et al., 1998]

Presently with the latest re-annotation of the M. tuberculosis genome, it is possible to assign a function to 2058 protein coding sequences (52% of the 3995 proteins predicted). Only 376 putative proteins share no homology with known proteins and thus could be unique to M. tuberculosis.

Drug resistance in Mycobacterium tuberculosis

Today eight agents are used in the treatment of TB. Multi drug resistant M. tuberculosis strains are resistant both to isoniazid (INH) and rifampicin (RFP). These agents have the most effective bactericidal activity towards M. tuberculosis. Nearly 95% of the RFP resistant strains possess a mutation in the rpoB gene encoding a DNA-dependent RNA polymerase. Approximately 90% of INH resistant strains have a mutation in the inhA, katG, and ahpG genes encoding enzymes related to a mycolic acid synthesis of cell wall. Pyrazinamide (PZA) resistant strains have a mutation in the pncA gene encoding a pyrazinamidase which degrades PZA to the bactericidal substance, pyrazinoic acid. Streptomycin resistant strains have a mutation in the rrs and rpsL gene encoding the 16S rRNA and the 12S ribosomal subunit protein, respectively. Kanamycin resistance is due to nucleotide substitutions in the rrs gene encoding 16S rRNA. Ethambutol resistant strains have a mutation in the embB gene encoding a arabinosyl transferase which catalyzes cell wall synthesis. Ethambutol resistance is in approximately 60% of organisms due to amino acid replacements at position 306 of an arabinosyltransferase encoded by the embB gene.

CLC bio provide some important bioinformatics solutions for detection and identification of mutations in selected genes of new strains of M. tuberculosis and other important pathogenic microbes. An example of this is described below.

Detection of mutations in Mycobacterium tuberculosis embB, rrs and pncA genes

The work flow for detecting and identifying mutations using the CLC Combined Workbench is summarized in figure 2.

Figure 2: Illustration of work flow.

The multi resistance M. tuberculosis strains are identified and the genomic DNA extracted. The encoding region of the embB, rrs and pncA genes are marked with primers using the graphical primer design. The relevant regions are sequenced automatically and assembled into contigs. BLAST searches in selected databases (both local and NCBI databases) are performed. Subsequently homologous sequences are aligned and annotations are transferred to the contig sequence.

In the next two sections, two of the steps in the work flow are described in further detail.

Zooming in on Primer design

Using the integrated GenBank search function, the annotated M. tuberculosis is downloaded. In a few clicks, the relevant genes (embB, rrs and pncA) are extracted. Guided by the CDS annotations, a set of primers are calculated for each gene (see figure 3). Using the interactive primer design functionality of the CLC Combined Workbench, the primers can be designed to match very specific needs.

Figure 3: Primer design guided by CDS annotations (the yellow boxes).

Zooming in on Assembly

The sequencing data is imported into the CLC Workbench, and during the assembly, the genes are automatically divided into separate contigs.

The inconsistencies that exist between different reads are inspected using the variance table (see figure 4).

Figure 4: Assembling of raw sequence data. The upper image is the contig variance table, showing where there are conflicts/variation between the reads. When selecting a row in the table, the bottom view automatically selects this position in the contig, and it can be inspected in more detail.

The reads are both forward and reverse, which are automatically detected during the assembly. The orientation is reflected in the color of the reads (in figure 4, the read at the top is reversed which is indicated by the red color of the residues). Quality scores are assigned in order to trim low-quality trace data, so that reads align properly. The quality scores can also be shown graphically below the sequence.

The result of the assembly is a contig sequence which is used in the following steps for BLAST searches and alignments.

The description above is only a brief example of how bioinformatics software, such as the CLC Workbenches, can be used in clinical microbiology. Although remarkable advances have been made, much remains to be learned about the molecular genetic basis of drug resistance development including M. tuberculosis. New therapeutics will be developed based on improved bioinformatic data-mining of results from bacterial drug resistance studies.