Cytoskeletons: Not Just for Eukaryotes

By: Sarah McKay, Natalie Stile, and Lauren Vaccaro


Ever since first learning about bacteria in elementary or middle school, there has been this instilled preconception that they are inherently inferior to eukaryotes. They evolved first, so they must be more primitive-- they’re smaller, lack organelles, and are generally unicellular. Quite the opposite from us. And because they are more primitive, they must be far less interesting than the multicellular organisms that we see every day, right?

In spite of misconceptions instilled in our early education, bacteria are complex and diverse, far more so than they may appear through casual observation. And as we’ve gotten better at studying both their structures and genomes, we’ve found many elements within them previously thought to be exclusive to eukaryotes. For example, the cytoskeleton: a meshwork of different protein polymers used to organize the cytoplasm and support cellular structure. The scientific community originally thought that this was another element unique to eukaryotes. Now, we have found many different types of cytoskeletal elements in prokaryotes that may not share the similar genetic codes, but share similar structures and functions. They tend to be classified in terms of the eukaryotic elements they resemble: actin homologs like MreB, which determines shape in rod-shaped cells, tubulin homologs like FtsZ, which polymerizes at the site of cell division, and intermediate filament-like elements, which can be quite varied.

Two examples of cytoskeletal elements produced by bacteria, both presented in vitro, compared to a
eukaryotic intermediate filament, keratin (left). The middle image is crescentin, an intermediate filament-
like protein that determines the shape of vibrioid bacteria like Caulobacter crescentus. The right is an
image of a cytoplasmic filamentous ribbon from Treponema phagedenis, a spirochetal bacteria. As you
can see, the prokaryotic and eukaryotic cytoskeletons are comparable. Images sourced from Lee and
Coulombe, 2009, Ausmees, 2006 and Izard, 2006.

One of the microbes scientists use to study cytoskeletal filaments is Streptomyces. If microbiology was a professional sport, Streptomyces would be the championship team - not only do they produce all kinds of pharmacologically relevant metabolites and antibiotics, they star in studies of microbe multi-cellularity, differentiation, and sporulation.

Of this star-studded crew, one species, Streptomyces coelicolor, stands out as an MVP. This model organism among model organisms is the most studied species of Streptomyces. It’s responsible for the production of over two thirds of all known natural antibiotics, but that’s not all! A soil bacteria that grows in networks of mycelia, S. coelicolor is a first-rate model for apical growth studies. Groups of S. coelicolor cells grow long, filamentous structures that extend from a specific pole known as the apical region. But how do they know to grow in one place and not another? A group of scientists, under the direction of Katsuya Fuchino, postulated that apical growth was somehow connected to the cytoskeleton (like many other microbial shape idiosyncrasies). Their recent paper, entitled Dynamic gradients of an intermediate filament-like cytoskeleton are recruited by a polarity landmark during apical growth, explores the relationship between cytoskeletal function and apical growth.

The morphology of S. coelicolor colonies.
As indicated by the name, it is quite colorful!
Source

From previous experimentation, Fuchino’s team already knew that the individual units that make up the cytoskeleton would randomly oligomerize, or come together in short chains, in vitro. Now the researchers wanted to move into the realm of living cells to investigate more specifically how the cytoskeletal proteins operate in vivo. Using a variety of different imaging and analysis techniques, Fuchino’s team observed a unique relationship between the two cytoskeletal proteins FilP and DivIVA, which operate at the apical region of S. coelicolor.

Of course, it’s unfair to discuss the intricacies of a novel biological process without first giving proper credit where it’s due. We proudly present FilP and DivIVA: two bacterial cytoskeletal proteins with promising biological research careers ahead of them. FilP is an intermediate filament-like, or “IF-like,” protein. A quick glance at its structure is quite revealing. FilP is a fibrous protein that has a coiled-coil domain architecture, very much like eukaryotic IF proteins. A coiled-coil is a protein secondary structure composed of two (or more) entwining alpha-helices that form a long cord, much like a rope. It also spontaneously assembles in the cytosol. In the IF-protein game, spontaneous assembly is like a prized skill on a resume: useful, efficient, and timeless. FilP requires no cofactors, nucleotides, or enzymes to facilitate its assembly. It’s all in the alpha-helices.

A rope, much like a coiled-coil, is composed of several
strands that are twisted together, forming a structure
stronger than that of the singular strands alone. This
makes coiled-coils ideal for cytoskeletal elements,
which are often used for supporting cellular structure.
Source

In microbial research, the prokaryotic FilP has a one-up on its eukaryotically-bound counterparts. Namely, that IF protein studies are difficult to carry out in eukaryotic cells. Numerous IF protein types exist across many, many kinds of differentiated adult cells, but few in this category (if any) are a part of tractable genetic systems. The molecular study of IF proteins is limited as a result. FilP, however, belongs to a prokaryote that fulfills these requirements! It’s a protein ripe for study, especially when knowing that a lot of its functionality is a mystery.

FilP, however, can’t function alone. It requires some guidance from its cooperative co-star, DivIVA. Like FilP, DivIVA is a coiled-coil protein. Unlike FilP, it has the unique feature of generating cell polarity. Cell polarity has a variety of biological uses. Need to know which direction to grow in? Polarity. Need to reach some resources? Polarity. Need to dive deeper into pond sludge for nutrients? Polarity. DivIVA helps simplify the process. At a cell’s pole formation sites, DivIVA assembles into a polarity-determining complex called a polarisome. Afterwards, the polarisome recruits the necessary building-blocks for cell envelope synthesis.

As important as background information on FilP and DivIVA is, it doesn’t answer our biggest (and most exciting) question: just what did Fuchino’s team discover, and why does it matter? We now know that DivIVA polarisomes “attract” FilP to S. coelicolor’s hyphal tips. and that FilP creates a concentration gradient in doing so. Logically, we could postulate that FilP and DivIVA were somehow interacting for this to occur. This is true, according to Fuchino’s results, but FilP and DivIVA (curiously) don’t co-localize. Instead, FilP assembles into a flexible, stress-bearing structure (similar to a volleyball net) that lets the hyphal tip grow. Thus, hyphal growth for S. coelicolor is stimulated in part by FilP and DivIVA’s strange interaction.

When allowed to polymerize in conditions similar
to those found in the cytosol, FilP forms a strong
hexagonal mesh. These structures help it remain
strong but flexible, like a volleyball net. Images
sourced from Fuchino et al. 2013 and here.

Fuchino and others  performed a number of experiments to analyze the assembly FilP, localization of DivIVA, and their interactive behavior. Firstly, they wanted to know how FilP assembled in vitro. The researchers found that FilP could form long, compact filaments but also built fibrous networks. These networks formed when FilP was incubated in a unique chemical buffer, Polymix, that resembled the cell cytosol. Secondly, Fuchino and his team used immunofluorescence to observe where FilP formed these networks during normal cell growth. FilP was strongly expressed in the apical tips of in Streptomyces hyphae. Moreover, it formed a concentration gradient which gradually weakened the farther it extended from growing hyphae. Surprisingly, despite their interactions, FilP and DivIVA did not colocalize, but were instead in close cellular proximity in young, branching hyphae.

At this point, Fuchino’s team had established that FilP assembles randomly in vitro and is localized to the apical region in vivo. Such a finding meant something else in the cell was causing FilP to assemble only at the apical region, and therefore Fuchino’s team further hypothesized that DivIVA was one of those factors. In another experiment, Fuchino’s team observed the localization of the two proteins in a strain of S. coelicolor which over expressed DivIVA along the sides of the cell as well as in the apical region. By placing DivIA in parts of the cell where it wouldn’t normally localize, it was possible to test the relationship between DivIVA and FilP. If DivIVA was involved with the recruitment of FilP, FilP would be visible at all the new points of DivIVA localization on the cell, including the apical region. If DivIVA wasn’t involved, FilP would continue to localize in the apical region, but not at the new areas of DivIVA expression. Fuchino’s team attached fluorescent signals to both proteins so their localization patterns would be easily viewed and tracked when DivIVA was overexpressed. What they found was that in 95% of the clusters of DivIVA, FilP was also present, supporting their hypothesis.

Above: An experiment run by Fuchino et al. (2013) showing that the localization of DivIVA

influences the localization of FilP. In this figure, green shows DivIVA, red shows FilP, and blue

shows the cell wall. In this strain, DivIVA is expressed not only at the apical region, but along

the sides as well. Notice that FilP is localized close to, but not in the same area as, DivIVA.

Below: Like players on the same team but in different positions, FilP and DivIVA work on the

same goal (apical growth) while being localized in different places. Image sourced from here.

The researchers also found that even though both FilP and DivIVA appeared in similar areas, they rarely overlapped in localization, which led Fuchino and cohorts to form a new hypothesis--DivIVA helps recruit FilP to the apical region, but the two do not co-localize. Like defensive and offensive players on the same team, the two proteins work together to protect the same goal, but they’re responsible for slightly different areas on the playing field.

While we still don’t truly understand their localization mechanisms, simply learning that FilP and DivIVA are interacting is a huge step forward. Thanks to this study, we know that not only do bacteria have cytoskeletons, but those cytoskeletal elements are interacting in ways not unlike those seen in eukaryotic cells. Further studies will be required to find the mechanism of their interactions, as well as how other molecules, like Scy, another coiled-coil protein believed to be involved, regulate and facilitate these interactions. While there is still much to learn about bacterial cytoskeletons and FilP in particular, we now have both a better groundwork for future studies and a greater appreciation for the complexity of our bacterial counterparts.



References

Ausmees, N. (2006). Intermediate filament-like cytoskeleton of caulobacter crescentus. Journal of Molecular Microbiology and Biotechnology, 11(3-5), 152-158.

Fuchino, K., Bagchi, S., Cantlay, S., Sandblad, L., Wu, D., Bergman, J., Kamali-Moghaddam, M., Flärdh, K., and Ausmees, N. (2013). Dynamic gradients of an intermediate filament-like cytoskeleton are recruited by a polarity landmark during apical growth. PNAS, 110(21), 1889-1897.

Izard, J. (2006). Cytoskeletal cytoplasmic filament ribbon of treponema: A member of an intermediate-like filament protein family. Journal of Molecular Microbiology & Biotechnology, 11(3-5), 159-166.

Lee, C., & Coulombe, P. A. (2009). Self-organization of keratin intermediate filaments into cross-linked networks. Journal of Cell Biology, 186(3), 409-421.