How do eukaryotic cells regulate gene expression

Each sequence of three bases is called a codon, and each codon contains the instructions for one amino acid. The tRNA assembles the protein using the amino acids. The protein continues to be built until a stop codon is encountered. A stop codon is a three-base sequence that does not code an amino acid. Three different types of RNA molecules are required for transcription and translation, and each type of RNA has a different function.

Each codon specifies a particular amino acid the building blocks of a protein. When the ribosome attaches to the mRNA, the codons are read. The start and stop codons do not code for any of the 20 amino acids. Proteins are the result of DNA transcription and translation.

Proteins are macromolecules made of one or more polypeptide chains, which are made up of a sequence of 20 different amino acids. Proteins bind to other molecules called ligands. After polypeptide chain s are completed, the chain s fold over onto themselves to create a 3-dimensional structure.

The resulting polypeptide chains direct the function of the protein in the cell. Antibodies bind to foreign particles eg, viruses, bacteria to protect the body. White blood cells, specifically B lymphocytes, produce antibodies. Enzymes perform or catalyze chemical reactions in cells eg, muscle contraction. Enzymes assist with bodily functions such as digestion and DNA replication.

Messenger proteins transmit signals to coordinate biological processes that occur between different cells, tissues and organs. Hormones eg, insulin, oxytocin are great examples of messenger proteins. Structural proteins provide structure and support for cells. Actin filaments and microtubules are examples of structural proteins. There are three types of structural proteins: fibrous, globular and membrane. Fibrous proteins form hair, nails and skin.

Transport proteins bind and carry atoms and small molecules. Hemoglobin is a transport protein in red blood cells that is used to carry oxygen from the lungs to other tissues. Because proteins are critical to supporting human life, it is imperative that they function properly. These data taken as a whole clearly suggest that molecular turnover at nuclear foci of Mig1 bound to its target genes occurs in units of single clusters, as opposed to single Mig1 monomers.

A Zoom-in on pairwise difference distribution for stoichiometry of Mig1-GFP foci, 7-mer intervals dashed and power spectrum inset , mean and Gaussian sigma error arrow. Inset shows the full range while outer zooms in on cluster stoichiometry. C 3C model of chromosomal DNA blue shaded differently for each chromosome with overlaid Mig1 promoter binding sites from bioinformatics red , simulated image based on model with realistic signal and noise added inset.

Our observations from stoichiometry, dynamics and kinetics, which supported the hypothesis that functional clusters of Mig1 perform the role of gene regulation, also suggested an obvious prediction in terms of the size of observed foci: the physical diameter of a multimeric cluster should be larger than that of a single Mig1 monomer.

We therefore sought to quantify foci widths from Slimfield data by performing intensity profile analysis on background-corrected pixel values over each foci image.

The diameter was estimated from the measured width corrected for motion blur due to particle diffusion in the sampling time of a single image frame, minus that measured from single purified GFP molecules immobilized to the coverslip surface in separate in vitro experiments. Immuno-gold electron microscopy of fixed cells probed with anti-GFP antibody confirmed the presence of GFP in 90 nm cryosections with some evidence of clusters containing up to 7 Mig1 molecules Figure 4—figure supplement 1C , however, the overall labeling efficiency was relatively low with sparse labelling in the nucleus in particular, possibly as a consequence of probe inaccessibility, resulting in relatively poor statistics.

Since we observed Mig1 clusters in live cells using Slimfield imaging we wondered if these could be detected and further quantified using other methods. Slimfield imaging on purified Mig1-GFP in vitro under identical imaging conditions for live cells similarly indicated monomeric Mig1-GFP foci in addition to a small fraction of brighter foci which were consistent with predicted random overlap of monomer images.

Depletion is an entropic derived attractive force which results from osmotic pressure between particles suspended in solution that are separated by distances short enough to exclude other surrounding smaller particles. These results support a hypothesis that clusters are present in live cells regardless of the concentration of extracellular glucose, which are stabilized by depletion components that are lost during biochemical purification.

We speculated that Mig1 cluster-mediated gene regulation had testable predictions in regards to the nuclear location of Mig1 at elevated extracellular glucose. We used sequence analysis to infer locations of Mig1 binding sites in the yeast genome, based on alignment matches to previously identified 17 bp Mig1 target patterns Lundin et al.

In scanning the entire S. We mapped these candidate binding sites onto specific 3D locations Figure 4C obtained from a consensus structure for budding yeast chromosomes based on 3C data Duan et al.

We generated simulated images, adding experimentally realistic levels of signal and noise, and ran these synthetic data through the same tracking software as for experimental data. We used identical algorithm parameters throughout and compared these predictions to the measured experimental stoichiometry distributions.

List of S. In the first instance we used these locations as coordinates for Mig1 monomer binding, assuming that just a single Mig1 molecule binds to a target. We assigned molecules to target promoter binding sites, then assigned the remaining 78 molecules randomly to non-specific DNA coordinates of the chromosomal structure.

We included the effects of different orientations of the chromosomal structure relative to the camera by generating simulations from different projections and included these in compiled synthetic datasets. We then contrasted monomer binding to a cluster binding model, which assumed that a whole cluster comprising 7 GFP labeled Mig1 molecules binds a single Mig1 target.

Here we randomly assigned the Mig1 molecules into just 27 i. We also implemented improvements of both monomer and cluster binding models to account for the presence of trans-nuclear tracks. We simulated the presence of these trans-nuclear molecules either using GFP-labeled Mig1 molecules as monomers, or as 18 i. This structural model supports the hypothesis that the functional unit of Mig1-mediated gene regulation is a cluster of Mig1 molecules, as opposed to Mig1 acting as a monomer.

We wondered if the discovery of transcription factor clusters was unique to specific properties of the Mig1 repressor, as opposed to being a more general feature of other Zn finger transcription factors.

To address this question we prepared a genomically encoded GFP fusion construct of a similar protein Msn2. Nrd1-mCherry was again used as a nuclear marker Figure 1—figure supplement 1. Msn2 acts as an activator and not a repressor, which co-regulates several Mig1 target genes but with the opposite nuclear localization response to glucose Lin et al.

On performing Slimfield under identical conditions to the Mig1-GFP strain we again observed a significant population of fluorescent Msn2 foci, which had comparable D and stoichiometry to those estimated earlier for Mig1 Table 2. These results support the hypothesis that two different eukaryotic transcription factors that have antagonist effects on the same target genes operate as molecular clusters. To test the functional relevance of Mig1 and Msn2 clusters we performed Slimfield on a strain in which Mig1 and Msn2 were genomically labeled using mCherry and orange fluorescent protein mKO2, respectively Lin et al.

This strain also contained a plasmid with GFP labeled PP7 protein to report on nuclear mRNA expressed specifically from the glycogen synthase GSY1 gene, whose expression can be induced by glucose starvation and is a target of Mig1 and Msn2, labelled with 24 repeats of the PP7 binding sequence Unnikrishnan et al.

No accumulation was observed with the mutant Mig1 lacking the Zn finger, in line with previous observations Lin et al. These results demonstrate a functional link between the localization of Mig1 and Msn2 clusters, and the transcribed mRNA from their target genes.

Since both Mig1 and Msn2 demonstrate significant populations of clustered molecules in functional cell strains we asked the question if there were features common to the sequences of both proteins which might explain this behavior.

To address this question we used multiple sequence alignment to determine conserved structural features of both proteins, and secondary structure prediction tools with disorder prediction algorithms.

As expected, sequence alignment indicated the presence of the Zn finger motif in both proteins, with secondary structure predictions suggesting relatively elongated structures Figure 6A.

We measured a trend from a more structured region of Mig1 towards the N-terminus and more disordered regions towards the C-terminus. Msn2 demonstrated a similar bipolar trend but with the structured Zn finger motif towards the C-terminus and the disordered sequences towards the N-terminus. We then ran the same analysis as a comparison against the prokaryotic transcription factor LacI, which represses expression from genes of the lac operon as part of the prokaryotic glucose sensing pathway.

An important observation reported previously is that the comparatively highly structured LacI exhibits no obvious clustering behavior from similar high-speed fluorescence microscopy tracking on live bacteria Mahmutovic et al. Intrinsically disordered proteins are known to undergo phase transitions which may enable cluster formation and increase the likelihood of binding to nucleic acids Uversky et al.

It has been shown that homo-oligomerization is energetically more favorable than hetero-oligomerization Goodsell and Olson, Moreover, symmetrical arrangement of the same protein can increase accessibility of the protein to binding partners, generate new binding sites, or increase complex specificity and diversity in general Fong et al.

We measured significant changes in circular dichroism of the Mig1 fusion construct upon addition of PEG in the wavelength range — nm Figure 6C known to be sensitive to transitions between ordered and intrinsically disordered states Sode et al. Since the Zn finger motif lies towards the opposite terminus to the disordered content for both Mig1 and Msn2 this may suggest a molecular bipolarity which could stabilize a cluster core while exposing Zn fingers on the surface enabling interaction with accessible DNA.

This structural mechanism has analogies to that of phospholipid interactions driving micelle formation, however mediated here through disordered sequence interactions as opposed to hydrophobic forces Figure 6C. The prevalence of phosphorylation sites located in disordered regions may also suggest a role in mediating affinity to target genes, similar to protein-protein binding by phosphorylation and intrinsic disorder coupling Nishi et al.

Figures and Tables. Predictions for the presence of intrinsically disordered sequences in Mig1, Msn2 and LacI, and of the positions of phosphorylation sites in Mig1 and Msn2. Our findings address a totally underexplored and novel aspect of gene regulation with technology that has not been available until recently. In summary, we observe that the repressor protein Mig1 forms clusters which, upon extracellular glucose detection, localize dynamically from the cytoplasm to bind to locations consistent with promoter sequences of its target genes.

Similar localization events were observed for the activator Msn2 under glucose limiting conditions. Our results therefore strongly support a functional link between Mig1 and Msn2 transcription factor clusters and target gene expression. The physiological role of multivalent transcription factor clusters has been elucidated through simulations Schmidt et al. These simulations show that intersegmental transfer between sections of nuclear DNA was essential for factors to find their binding sites within physiologically relevant timescales and requires multivalency.

Previous single-molecule studies of p53 Mazza et al. Our findings address the longstanding question of how transcription factors find their targets in the genome so efficiently. Evidence for higher molecular weight Mig1 states from biochemical studies has been suggested previously Needham and Trumbly, A Mig1-His-HA construct was overexpressed in yeast and cell extracts run in different glucose concentrations through sucrose density centrifugation.

In western blots, a higher molecular weight band was observed, attributed to a hypothetical cofactor protein. However, no cofactor was detected and none reported to date. The authors only reported detecting higher molecular weight states in the nucleus in repressing conditions. Clustering of nuclear factors has been reported previously in other systems using single-molecule techniques.

In particular, RNA polymerase clustering in the nucleus has been shown to have a functional role in gene regulation through putative transcription factories Cisse et al. Other nuclear protein clusters have been shown to have a functional role Qian et al. Our measured turnover of genome-bound Mig1 has similar timescales to that estimated for nucleoid-bound LacI Mahmutovic et al.

Faster off rates have been observed during single particle tracking of the DNA-bound fraction of the glucocorticoid receptor GR transcription factor in mammalian cells, equivalent to a residence time on DNA of just 1 s Gebhardt et al.

Single GR molecules appear to bind as a homodimer complex on DNA, and slower Mig1 off rates may suggest higher order multivalency, consistent with Mig1 clusters.

Estimating nearest-neighbor distances between Mig1 promoter sites in the S. Such multivalency chimes with the tetrameric binding of prokaryotic LacI leading to similar low promoter off rates Mahmutovic et al. Measuring the variation of electrostatic charge of residues for the amino acid sequences of both Mig1 and Msn2 Figure 6F we see that the regions in the vicinity of the Zn finger motifs for both proteins have a strong net positive charge compared to the rest of the molecule.

If these regions project outwards from a multivalent transcription factor cluster, as per our hypothesized cluster model Figure 6E , then the cluster surface could interact electrostatically with the negatively charged phosphate backbone of DNA to enable a 1D sliding diffusion of the protein along a DNA strand, such that the on rate for the protein-DNA interaction is largely sequence-independent in regards to the DNA.

Particular details of this type of transcription factor binding to non-specific regions of DNA have been investigated at the level of single transcription factor molecules using computational simulations Rohs et al. This lack of sequence dependence for binding is consistent with observations from an earlier live cell single-molecule tracking study of the TetR repressor Normanno et al.

We also see experimental evidence for this in our study here, in that we find that the best fit model to account for fluorescence images of the nucleus under high glucose conditions is a combination of occupancy of clusters at the target genes i.

Ultimate binding to the gene target once encountered could then be mediated through sequence-specific interactions via the Zn finger motif itself. If the haploid genome of budding yeast, containing This tube diameter, in the absence of local contributions from histone packing, is thus a rough estimate for the effective average separation of DNA strands in the nucleus i. A multivalent transcription factor cluster thus may have only a relatively short distance to diffuse through the nucleoplasm if it dissociates from one DNA strand and then rebinds electrostatically to the next nearest strand, thereby facilitating intersegmental transfer.

In this scheme, the association interaction between clusters and neighboring DNA strands is predominantly electrostatic and therefore largely, one might speculate, sequence-independent. However, sequence specificity may be relevant in generating higher-order packed structures of chromatin resulting in localized differences to the nearest neighbor separation of different DNA strands, which could therefore influence the rate at which a cluster transfers from one strand to another.

In addition, there may also be localized effects of DNA topology that affect transcription factor binding, which in turn would be expected to have some sequence specificity Rohs et al. Also, the off rates of cluster interactions with DNA may be more dependent on the specific sequence. For example, one might anticipate that the dissociation of translocating clusters would be influenced by the presence of obstacles, such as other proteins, already bound to DNA which in turn may have sequence specificity.

In particular, bound RNA polymerases present during gene transcription at sequence specific sites could act as roadblocks to kick off translocating clusters from a DNA strand, to again facilitate intersegmental transfer.

Several previous experimental studies report observations consistent with intersegmental transfer relevant to our study here. For example, an investigation using single-molecule tracking indicated that transcription factor search times were increased if intersegmental transfer was specifically abrogated Elf et al.

These observations are consistent with other experiments that selectively enabled intersegmental transfer by altering DNA conformation Lomholt et al. Also, they are consistent with biochemical measurements that transcription factors spend a high fraction of their time bound to DNA, as opposed to being in solution Elf et al. Furthermore, other light microscopy studies report direct experimental evidence for intersegmental transfer Gowers and Halford, ; Gowers et al.

It is well-established from multiple studies that 3D diffusion of transcription factors in the nucleoplasm alone cannot account for the relatively rapid search times observed experimentally to find specific targets in the genome Berg et al. Constraining the dimensionality of diffusion to just 1D, as in the sliding of weakly bound transcription factors on DNA, speeds up this process, but is limited by encountering obstacles already bound to the DNA which potentially result in dissociation of the transcription factor and then slow 3D diffusion in the nucleoplasm.

In our system, we speculate that the clusters we observe can slide on DNA in a largely sequence-independent manner but then can cross to neighboring DNA strands in a process likely to have some sequence dependence when an obstacle is encountered, and thus predominantly circumvent the requirement for slow 3D diffusion in the nucleoplasm.

Minimizing the contribution from the slowest component in the search process may therefore result in an overall reduction in the amount of time required for a given transcription factor to find its gene target.

Extensive bioinformatics analysis of proteome disorder across a range of species suggests a sharp increase from prokaryotes to eukaryotes Xue et al.

Our discovery in yeast may reveal a eukaryotic adaptation that stabilizes gene expression. The slow off rate we measure would result in insensitivity to high frequency stochastic noise which could otherwise result in false positive detection and an associated wasteful expression response. Our results suggest that cellular depletion forces due to crowding enable cluster formation.

Crowding is known to increase oligomerization reaction rates for low association proteins but slow down fast reactions due to an associated decrease in diffusion rates, and have a more pronounced effect on higher order multimers rather than dimers Phillip and Schreiber, It is technically challenging to study depletion forces in vivo, however there is growing in vitro and in silico evidence of the importance of molecular crowding in cell biology.

A particularly striking effect was observed previously in the formation of clusters of the bacterial cell division protein FtsZ in the presence of two crowding proteins — hemoglobin and BSA Rivas et al. Similarly, two recent yeast studies of the high-osmolarity glycerol HOG pathway also suggest a dependence on gene expression mediated by molecular crowding Babazadeh et al.

The range of GFP labeled Mig1 cluster diameters in vivo of 15—50 nm is smaller than the 80 nm diameter of yeast nuclear pore complexes Ma and Yang, , not prohibitively large as to prevent intact clusters from translocating across the nuclear envelope. An earlier in vitro study using sucrose gradient centrifugation suggested a Stokes radius of 4.

The authors ascribed this effect to a hypothetical elongated monomeric structure for Mig1. The equivalent Stokes radius for GFP has been measured at 2. Thus the anticipated hydrodynamic diameter of Mig1-GFP is 15—16 nm. Using Stokes law this estimated hydrodynamic radius indicates an effective viscosity for the cytoplasm and nucleoplasm as low as cP, compatible with earlier live cell estimates on mammalian cells using fluorescence correlation spectroscopy FCS Liang et al.

One alternative hypothesis to that of intrinsically disordered sequences mediating Mig1 cluster formation is the existence of a hypothetical cofactor protein to Mig1. However, such a cofactor would be invisible on our Slimfield assay but would result in a larger measured hydrodynamic radius than we estimate from fluorescence imaging, which would be manifest as larger apparent viscosity values than those we observe. Coupled to observations of Msn2 forming clusters also, and the lack of any reported stable cofactor candidate to date, limits the cofactor hypothesis.

Pull down assays do suggest that promoter bound Mig1 consists of a complex which includes the accessory proteins Ssn6 and Tup1 Treitel and Carlson, , however this would not explain the observation of Mig1 clusters outside the nucleus. There may be other advantages in having a different strategy between S.

A clue to these may lie in phosphorylation. Thus phosphorylation at sites within these regions may potentially disrupt binding to DNA, similar to observed changes to protein-protein affinity being coupled to protein phosphorylation state Nishi et al. Wide scale bioinformatics screening reveals a significant prevalence of intrinsic disorder in eukaryotic transcription factors Liu et al. Our discovery is the first, to our knowledge, to make a link between predicted disorder and the ability to form higher-order clusters in transcription factors.

Thus, our results address the longstanding question of why there is so much predicted disorder in eukaryote transcription factors. Our observations that protein interactions based on weak intracellular forces and molecular crowding has direct functional relevance may stimulate new research lines in several areas of cell biology. Increased understanding of the clustering mechanism may not only be of value in understanding such diseases, but could enable future novel synthetic biology applications to manufacture gene circuits with, for example, a range of bespoke response times.

All transformations were performed using the lithium acetate protocol Gietz and Schiestl, Cell doubling times of all strains were calculated Warringer et al. Extracts were cleared 24, g, 30 min and filtered pore diameter 0.

Mig1-mGFP was eluted with a linear gradient 0—0. OD measurements at nm were taken every 10 min with prior shaking. Each strain was represented in sextuplicates. Images were acquired not longer than 2 hr after the last media switch.

The following procedure was adapted from Bendrioua et al Bendrioua et al. Cells were harvested by centrifugation 3, rpm, 50 s , suspended in 1 ml of 0. Samples were vortexed and spun down at 13, rpm. The pellets were washed in 0. The protein extracts were obtained by centrifuging at the maximal speed and collecting the supernatants.

The cultures were harvested by centrifugation, suspended in 0. A dual-color bespoke laser excitation single-molecule fluorescence microscope was used Badrinarayanan et al. Fluorescence emissions were captured by a 1. We confirmed negligible measured crosstalk between GFP and mCherry signals to red and green channels respectively, using purified GFP and mCherry sampled in an in vitro surface immobilization assay details below. Three color microscopy was performed on the same microscope, using a 50 mW nm wavelength laser Obis to excite mKO2, coupled into the same optics as before with the addition of a nm notch rejection filter Semrock, Rochester, New York, UK in both channels of the imaging path.

This allowed 1 mW of laser excitation at the sample. To investigate time-resolved glucose concentration-dependent changes in Mig1-GFP localization in individual yeast cells, we used bespoke microfluidics and our bespoke control software CellBild LabVIEW, National Instruments, Austin, Texas, United States , enabling cell-to-cell imaging in response to environmental glucose changes. CellBild controlled camera acquisition synchronized to flow-cell environmental switches via a syringe pump containing an alternate glucose environment.

Microfluidic flow-chambers were based on an earlier 4-channel design Gustavsson et al. Cells were introduced via a side channel and were left to bind ConA for 15 min to immobilize cells on the surface. Slimfield imaging was performed on a similar bespoke microscope setup at comparable laser excitation intensities and spectral filtering prior to imaging onto a Photometrics Evolve Delta EMCCD camera at frames per second.

Alternating frame laser excitation ALEX was used to minimize any autofluorescence contamination in the red channel introduced by the blue excitation light. Around 1—4 cells were imaged in a single field of view for each glucose exchange. The same flow chamber was used for multiple fields of view such that each cell analyzed may have experienced up to four glucose exchange cycles.

Our bespoke foci detection and tracking software objectively identifies candidate bright foci by a combination of pixel intensity thresholding and image transformation to yield bright pixel coordinates. The intensity centroid and characteristic intensity, defined as the sum of the pixel intensities inside a 5 pixel radius region of interest around the foci minus the local background and corrected for non-uniformity in the excitation field are determined by iterative Gaussian masking.

Foci in consecutive image frames within a single point spread function PSF width, and not different in brightness or sigma width by more than a factor of two, are linked into the same track. The microscopic diffusion coefficient D is then estimated for each accepted foci track using mean square displacement analysis, in addition to several other mobility parameters. Hoopes, L. Nature Education 1 1 Atavism: Embryology, Development and Evolution.

Gene Interaction and Disease. Genetic Control of Aging and Life Span. Genetic Imprinting and X Inactivation. Genetic Regulation of Cancer. Obesity, Epigenetics, and Gene Regulation. Chromatin Remodeling and DNase 1 Sensitivity.

Chromatin Remodeling in Eukaryotes. The Role of Methylation in Gene Expression. Gene Expression Regulates Cell Differentiation. DNA Transcription. Reading the Genetic Code. Simultaneous Gene Transcription and Translation in Bacteria.

Negative Transcription Regulation in Prokaryotes. Operons and Prokaryotic Gene Regulation. Regulation of Transcription and Gene Expression in Eukaryotes.

Environmental Influences on Gene Expression. Genes, Smoking, and Lung Cancer. RNA Functions. Explore This Subject. Consequences of Gene Regulation.

Organization of Chromatin. Transcription Factors. From DNA to Protein. Regulation of Transcription. Gene Responses to Environment. You have authorized LearnCasting of your reading list in Scitable. Do you want to LearnCast this session? This article has been posted to your Facebook page via Scitable LearnCast. Change LearnCast Settings.