24/04/25
23:30:59
Laboratory of Mathematic methods and models in bioinformatics,
Institute for Information Transmission Problems,
Russian Academy of Sciences

Classical Attenuation Regulation

Problem statement and method description

Often, the signal is characterized by additional requirements for the target sites. For example, one might require that the signal allows for a specific combination of helices arranged in a certain way relative to each other. A helix is defined as a pair of segments ("arms") that can be paired letter-by-letter according to the rule G with C and A with T. In this case, we speak of a signal based on secondary structure; the original sequences themselves are referred to as primary structures. Searching for such a signal, i.e., ultimately constructing a multiple alignment, presents a significant challenge. Here, the signal under consideration is the classical attenuation regulation, described, for example, in [1].

We have proposed a method based on an improvement of the model from [1]. Specifically, this model now accounts for RNA triplexes, pseudoknots of various types, and also premature transcription termination, which can occur not only at poly-U tracts but also anywhere along the RNA polymerase's path, provided the inhibitory effect of the corresponding helix unfavorably combines with the weakening of the polymerase's bond to the DNA. This model was used to select leader regions of amino acid genes potentially containing this regulation, followed by alignment of the regulatory sites. Thus, the selection of leader regions was fundamentally different from the approach in [2], where the described LLLM algorithm was used. This method was applied extensively to data from the NCBI database.

Results of method application

For the hisG gene across the spectrum, except for the Pseudomonadaceae and Xanthomonadaceae families of gamma-proteobacteria (the alignment at the provided link includes 64 leader regions of this gene simultaneously), four conserved helices are observed after the leader peptide gene. The first of these acts as an antiterminator; the antiterminator (helix 1) does not overlap with the terminator (helix 4) but transmits its effect through a co-antiterminator (helix 2, which forms in the presence of the antiterminator). If the antiterminator cannot form due to the ribosome's position on the leader peptide gene's mRNA, a co-terminator (helix 3) forms instead, preventing helix 2 from forming but allowing the terminator to form. As our model shows, the formation of triplexes on the antiterminator and partially on the co-terminator is essential for the regulation process.

A similar regulatory pattern for the genes ilvD, lysQ, hisZ, and trpB is observed in a smaller alignment among some firmicutes. The lysQ gene encodes a permease. It is possible that in the single species Lactococcus lactis, this permease has changed specificity, as we observe histidine concentration-dependent regulation, which is absent in other firmicutes. For the hisZ gene, such regulation is observed in bacilli, listeria, and a single clostridium, but in listeria, triplexes do not form.

For the trpB gene, regulation is predicted only in two firmicutes: Bacillus halodurans and Bacillus thuringiensis. Triplexes are not involved in the regulation, and the structure is poorly conserved. This is a rare case of this gene being regulated as the first in an operon; it is also regulated this way in a single actinobacterium, Corynebacterium diphtheriae. In gamma-proteobacteria, it is regulated as part of the long operon trpEGDCFBA.

For the ilvD gene, this regulation is observed in Staphylococcus and Listeria. However, for the ilvD gene in Geobacillus thermodenitrificans, classical regulation is observed, where the antiterminator directly overlaps with the terminator and no triplex forms. A preliminary result for this case was presented in [3].

For the genes hisG, ilvD, and trpE in Bacteroides across their spectrum, regulation in its traditional form—where the antiterminator directly overlaps with the terminator—was found in more than 48 cases.

For the genes hisS, trpE, and ilvD in Thermotogae and Chloroflexi, regulation was also found in several cases, where the antiterminator directly overlaps with the terminator.

References

  1. A.V. Seliverstov, V.A. Lyubetsky. Translation regulation of intron-containing genes in chloroplasts. Journal of Bioinformatics and Computational Biology, 2006, Vol. 4, No. 4, P. 783–792. DOI: 10.1142/S0219720006002235
  2. A.G. Vitreschak, E.V. Lyubetskaya, M.A. Shirshin, M.S. Gelfand, V.A. Lyubetsky. Attenuation regulation of amino acid biosynthetic operons in proteobacteria: comparative genomics analysis. FEMS Microbiology Letters, 2004, Vol. 234, Iss. 2, P. 357–370. DOI: 10.1016/j.femsle.2004.04.005
  3. V.A. Lyubetsky, A.V. Seliverstov. Amino acid biosynthesis attenuation in bacteria. Proceedings of the Fourth International Conference on Bioinformatics of Genome Regulation and Structure (BGRS'2004), Novosibirsk, Russia, July 25–30 2004, Vol. 1, P. 307–310.