A simple regulatory architecture allows learning the statistical structure of a changing environment

Bacteria live in environments that are continuously fluctuating and changing. Exploiting any predictability of such fluctuations can lead to an increased fitness. On longer timescales, bacteria can ‘learn’ the structure of these fluctuations through evolution. However, on shorter timescales, inferring the statistics of the environment and acting upon this information would need to be accomplished by physiological mechanisms. Here, we use a model of metabolism to show that a simple generalization of a common regulatory motif (end-product inhibition) is sufficient both for learning continuous-valued features of the statistical structure of the environment and for translating this information into predictive behavior; moreover, it accomplishes these tasks near-optimally. We discuss plausible genetic circuits that could instantiate the mechanism we describe, including one similar to the architecture of two-component signaling, and argue that the key ingredients required for such predictive behavior are readily accessible to bacteria.

Genetically Engineered Integral Feedback Controllers for Robust Perfect Adaptation in Mammalian Cells

Mammalian cells collectively maintain a consistent internal milieu that supports their host’s survival in varying and uncertain environments. This homeostasis is often achieved through negative feedback loops that act at various levels of biological organization, from the system and organ levels down to gene expression at the molecular scale. Recently, a molecular regulatory motif has been discovered that enables a regulated variable to adapt perfectly (at the steady state) to network and parameter changes and to persistent environmental perturbations. The regulatory motif that achieves this robust perfect adaptation property realizes integral feedback, a control strategy that employs mathematical integration in a negative feedback loop. Here, we present the first synthetic implementation of integral feedback in mammalian cells. We show that this implementation successfully maintains constant levels of a transcription factor, even when its degradation is significantly increased. Furthermore, we establish the structural robustness properties of our controlled system by demonstrating that perturbing the network topology does not affect the transcription factor levels. We believe that the ability to robustly and predictably regulate the expression levels of genes will both become an indispensable tool for basic research as well as lead to substantial advances in the development of industrial biotechnology and cell-based therapies.

Assessing deep learning algorithms in cis-regulatory motif finding based on genomic sequencing data

Cis-regulatory motif finding is a crucial step in the detection of gene regulatory mechanisms using genomic data. Deep learning (DL) models have been utilized to denovoly identify motifs, and have been proven to outperform traditional methods. By 2020, twenty DL models have been developed to identify DNA and RNA motifs with diverse framework designs and implementation styles. Hence, it is beneficial to systematically compare their performances, which can facilitate researchers in selecting the appropriate tools for their motif analyses. Here, we carried out an in-depth assessment of the 20 models utilizing 1,043 genomic sequencing datasets, including 690 ENCODE ChIP-Seq, 126 cancer ChIP-Seq, 172 single-cell cleavages under targets and release using a nuclease, and 55 RNA CLIP-Seq. Four metrics were designed and investigated, including the accuracy of motif finding, the performance of DNA/RNA sequence classification, algorithm scalability, and tool usability. The assessment results demonstrated the high complementarity of the existing models, and it was determined that the most suitable model should primarily depend on the data size and type as well as the model outputs. A webserver was developed to allow efficient access of the identified motifs and effective utilization of high-performing DL models.

Erratum: Mohammad Shibli Kaysar and Mohammad Ibrahim Khan A Modified Median String Algorithm for Gene Regulatory Motif Classification Symmetry 2020, 12, 8

The authors wish to make the following corrections to their paper: [...]

A Modified Median String Algorithm for Gene Regulatory Motif Classification

Consensus string is a significant feature of a deoxyribonucleic acid (DNA) sequence. The median string is one of the most popular exact algorithms to find DNA consensus. A DNA sequence is represented using the alphabet Σ= {a, c, g, t}. The algorithm generates a set of all the 4l possible motifs or l-mers from the alphabet to search a motif of length l. Out of all possible l-mers, it finds the consensus. This algorithm guarantees to return the consensus but this is NP-complete and runtime increases with the increase in l-mer size. Using transitional probability from the Markov chain, the proposed algorithm symmetrically generates four subsets of l-mers. Each of the subsets contains a few l-mers starting with a particular letter. We used these reduced sets of l-mers instead of using 4ll-mers. The experimental result shows that the proposed algorithm produces a much lower number of l-mers and takes less time to execute. In the case of l-mer of length 7, the proposed system is 48 times faster than the median string algorithm. For l-mer of size 7, the proposed algorithm produces only 2.5% l-mer in comparison with the median string algorithm. While compared with the recently proposed voting algorithm, our proposed algorithm is found to be 4.4 times faster for a longer l-mer size like 9.

Tunable genetic devices through simultaneous control of transcription and translation

Synthetic genetic circuits allow us to modify the behavior of living cells. However, changes in environmental conditions and unforeseen interactions with the host cell can cause deviations from a desired function, resulting in the need for time-consuming reassembly to fix these issues. Here, we use a regulatory motif that controls transcription and translation to create genetic devices whose response functions can be dynamically tuned. This allows us, after construction, to shift the on and off states of a sensor by 4.5- and 28-fold, respectively, and modify genetic NOT and NOR logic gates to allow their transitions between states to be varied over a >6-fold range. In all cases, tuning leads to trade-offs in the fold-change and the ability to distinguish cellular states. This work lays the foundation for adaptive genetic circuits that can be tuned after their physical assembly to maintain functionality across diverse environments and design contexts.

An NAD+-dependent novel transcription factor controls stage conversion in Entamoeba

Developmental switching between life-cycle stages is a common feature among parasitic pathogens to facilitate disease transmission and pathogenesis. The protozoan parasite Entamoeba switches between invasive trophozoites and dormant cysts, but the encystation process remains poorly understood despite being central to amoebic biology. We identify a transcription factor, Encystation Regulatory Motif-Binding Protein (ERM-BP), that regulates encystation. Down-regulation of ERM-BP decreases encystation efficiency resulting in abnormal cysts with defective cyst walls. We demonstrate that direct binding of NAD+ to ERM-BP affects ERM-BP conformation and facilitates its binding to promoter DNA. Additionally, cellular NAD+ levels increase during encystation and exogenous NAD+ enhances encystation consistent with the role of carbon source depletion in triggering Entamoeba encystation. Furthermore, ERM-BP catalyzes conversion of nicotinamide to nicotinic acid, which might have second messenger effects on stage conversion. Our findings link the metabolic cofactors nicotinamide and NAD+ to transcriptional regulation via ERM-BP and provide the first mechanistic insights into Entamoeba encystation.