Say what?

Last week, a research article that I contributed to almost 15 years ago was finally published. The reasons for the long delay aren’t important (and I don’t know them anyway, lol), but this is one of those things that occasionally happens in the world of scientific research. Specific projects generally “belong” to graduate students or postdocs, which are scientists-in-training that work under the supervision of a PI (Principle Investigator, the faculty member or senior scientist in whose lab the work takes place). But people often move on from these positions after a few years and occasionally, a good project with solid results will languish for years for no other reason than no one gets around to finishing it. Labs are busy places; everyone has multiple projects already; grant funding comes and goes; and individual researchers move on to other projects or positions.

Here is the article:

(This project was mostly the work of Taihao Quan and Gary J. Fisher, both at the University of Michigan Medical School, but several others contributed, including me, as a collaborator.)

When I got the little “ping” from Google Scholar that a new article had been added to my profile, I got that warm feeling that I always get when I publish something, however small and technical, that adds to what we know about how life works. For scientists, our research articles represent the knowledge we discovered and contributed to the world.

As proud as I am of the books and articles I have written for the general public, my research articles hold a special place because they are my cred, which could equally mean “credibility” or “credentials,” because it’s how scientists tend to size each other up. So, I thought I would take a minute and describe the work that was just published, because it highlights part of my scientific life – molecular and cellular biology – that I don’t often blog about.

My work in molecular biology is still very much who I am as a scientist. The work that scientists do during training – graduate school and postdoctoral fellowships – becomes the core of our scientific identity. My PhD advisor, despite working in cell and molecular biology for more than three decades, will always think of himself as a biophysicist because that’s what he did for his PhD and his first postdoc. I have been working in the field of genome evolution for over ten years, but I still think of myself first as a cellular and molecular biologist. So I thought I would use this article as a way to give you all a glimpse into another side of my scientific life. I hope you enjoy it!

Okay, here we go. This project focused on dermal fibroblasts and how they reproduce themselves. As you might guess, dermal refers to the skin, but specifically to an underneath layer called the dermis. The outermost layer that people are most familiar with is called the epidermis (epi- means “on top of”) and is composed primarily of dead keratinocytes, which are chock full of keratin proteins. The epidermis is a water-tight barrier of oily, protein-packed cells that are dead so they cannot be infected by viruses or other invaders.

But the layer underneath, the dermis, is very much alive and has a rich blood supply. If you scratch yourself and do not bleed, you only scraped the epidermis. But if your scrape bleeds, it reached the dermis. (But hopefully not the subcutaneous layer (aka hypodermis), or else, off to the ER you go!). Unlike the epidermis, which is pretty much just cells connected to each other, the dermis has a lot of something called extracellular matrix. This is a meshwork of proteins that gives skin its shape, structure, strength, and flexibility. Dermal fibroblasts are the cells in the dermis that make the extracellular matrix. (Because extracellular is such a mouthful, we often just call it “the matrix.”)

Because living cells are basically just tiny bags of water, the matrix is what actually gives tissues and organs their shape. It’s like a skeleton, or a scaffold, or the frame of a house. It provides shape, sturdiness, and all of the physical structure of the organ or tissue. The matrix is like a sponge, and the cells are like the water held by a saturated sponge. Even if the water provides most of the mass of the saturated sponge, the sponge itself is what provides the shape. That’s like the matrix. It gives tissues and organs their shape, and the cells live inside.

The main protein of the matrix is collagen, and the matrix is such a big part of our bodies that collagen is the most abundant protein in the human body, by far. There are something like 20,000 different proteins in the human body (up to a million if you count alternative splice forms and post-translational modifications), but 30% of our total protein mass is collagen. That’s twice as much as the combined protein mass of all the muscle fibers in the body. Collagen is very much like a rope – a triple-braid of fibers, which are then coiled again and again to form very strong cables. Another important protein in the matrix is elastin, which, as you can probably guess, gives tissues their stretchiness and flexibility, their elasticity, in technical terms.

Different tissues have different proportions of the various types of collagen and elastin. Most skin, for example, is somewhat loosely stitched, with a lot of elastin, making it stretchy, but not all that strong. Tendons and ligaments, on the other hand, are made almost entirely of type I and type III collagen, which has a very high tensile strength, but is not very stretchy.

Okay, so now we are almost ready to dive in to the paper. As I mentioned, fibroblasts are the cells responsible for making and remodeling the extracellular matrix. They synthesize a lot of collagen and then organize it into a mesh. This also means that they do most of the work of healing wounds. Whenever a tissue is injured or disrupted, stopping the bleeding and fixing blood vessels is just one part of healing the injury. The tissue itself must be repaired and this means that fibroblasts must rebuild the matrix and replace the damaged parts. The skin is subjected to more minor injuries than any other tissue, and so dermal fibroblasts are the real work horses of wound healing. Scientists study dermal fibroblasts as a model for wound healing throughout the body.

Wound healing is not a perfect process. The most common complication is called fibrosis, which is the formation of a scar. In the skin, scars can be unsightly, but large ones can also cause medical problems. This is because scars are made of thick mats of collagen. They are not stretchy and springy, and so when they get pulled on, the tissue around the scar can tear, causing a further wound, which is then healed, causing more scar tissue to form, leading to further stretch-injuries, and so on. In the skin, this is usually no big deal, but when fibrosis occurs in other tissues, such as the lungs, liver, kidney, and heart, this can lead to life-threatening complications. Scars are not just ugly, they can impair the function of vital organs.

One of the proteins that is involved in wound healing and scar formation is called CCN2. (CCN stands for cell communication network, and CCN2 is also known as CTGF, connective tissue growth factor.) When a tissue is damaged, fibroblasts are attracted to the site of the injury. Once they arrive, they release CCN2 into the environment, which then stimulates other nearby fibroblasts to grow and multiply. This is an important function of CCN2 because in order to heal a wound, the matrix will need to be rebuilt. That takes a lot of collagen and a lot of fibroblasts. So fibroblasts that are hard at work on tissue repair release CCN2 as a way to make more of themselves. Increasing their number speeds the rate of healing, but it also leads to fibrosis – too many fibroblasts and too much collagen is what forms a scar.

In order to prevent and reverse scar formation, we need to fully understand how it happens. That was the motivation behind this research project. And the specific research question was, how does CCN2 tell fibroblasts to proliferate? Once we know that, it may be possible to promote or allow its function during normal wound healing, but inhibit or reverse it during scar formation.

As I said, scar formation is not just cosmetic (although the cosmetics industry is very interested in this research area). This project has the potential to contribute to reversing liver damage, kidney disease, lung injuries and pathologies (including emphysema/COPD), and damage to the heart muscle following a heart attack. All of those are at least partially caused by fibrosis, so the implications of this research are far-reaching. (The excitement of these possibilities is what motivates scientists to do our work!)

It was well known that when dermal fibroblasts are treated with CCN2 protein, they grow and proliferate. In the first experiment of our paper, we showed that if you remove CCN2, the opposite happens, the fibroblasts STOP proliferating. In the micrographs below (photographs taken with a microscope), CTRL si = the control, and CCN2 si = removal of CCN2. Note how, in the controls, the cells proliferate between day 2 and day 5 (panels 1 and 2). But when CCN2 has been removed, they do not (panels 3 and 4).

This experiment showed that CCN2 doesn’t just promote the proliferation of dermal fibroblasts, but that it is required for their proliferation, at least under the laboratory conditions that they are normally grown in. (It’s always possible that cells behave differently in the body than they do in lab experiments, but it’s not exactly ethical to do these experiments in living people, so this is the best we can do. Also, for simplicity, I’m skipping most of the control and duplication experiments, but rest assured that for each of our results, we repeated the experiments a few different ways for confirmation.)

The next question we asked was, when CCN2 is removed and the cells stop multiplying, in what phase of the cell cycle do they get stuck? The cell cycle is the orderly series of steps that cells take when they are proliferating. These steps are organized into phases. In G1 phase, the cells are mostly just growing larger, almost doubling in size. In S phase, the cells duplicate all of their DNA, so that when the cell splits in two, each cell will have the full set of genes. In G2 phase, cells scan for any errors or damage to the DNA that can result from all of that copying and they fix the damage if they can. (If they can’t, they undergo apoptosis, also known as programmed cell death or cell suicide, in order to protect our bodies from the dangers of a damaged cell.) And the final phase is M phase, or mitosis and cytokinesis, when the duplicated chromosomes are separated and the cell splits in two. Once we know which phase the fibroblasts get stuck in when CCN2 is removed, we may get a clue as to how it works to promote cell proliferation.

In the far right panel of the figure below, you can see how all of the cells are stuck in G1 phase. This looks like a subtle difference here because cells spend most of their time in G1 anyway, but in this experiment you can see that almost ALL of the cells (95%) are in G1.

Since removal of CCN2 causes cells to pile up in G1 phase of the cycle, we turned our attention to the various events that take place in G1 in order to narrow down precisely where they get stuck. There are two families of proteins that control how a cell progresses through G1 phase. The cyclins and the CDKs (cyclin-dependent kinases). When they are both present and active, they work together to push a cell forward.

CDKs alone are inactive and the cell is paused. CDK + cyclin are active and push the cell forward. Because the cells were paused in G1 phase, we started looking at the G1 phase cyclins and CDKs to see what might be causing the cells to pause.

There are two groups of G1 phase cyclins and CDKs. Early in G1 phase, the D cyclins accumulate and activate the specific CDKs CDK4 and CDK6. In late G1 phase, E cyclin appears and activates CDK2. (CDK2 is also important during S phase, but at that point, it partners with A cyclin, not E cyclin.) Both cyclin D-CDK4/6 and cyclin-CDK2 target a protein called the retinoblastoma protein (pRB) and add phosphates to it. pRB’s job is to hold the cells in early G1 phase, and when the cyclin-CDKs add phosphates to pRB, they inactive it, allowing the cells to move forward in the cycle.

So because our cells were accumulating in G1 phase when we removed CCN2, we needed to determine which cyclin was involved, D or E; as well as which CDK was involved CDK4/6 or CDK2. Below is the result of many experiments that we performed to analyze the status of the G1 cyclins and CDKs.

This would take a long time to fully describe, so I’ll give you the summary. Panel A shows that cyclin E does NOT accumulate when we remove CCN2. Panel B shows that CDK2 does NOT become active when we remove CCN2. These two panels tell us that the block in G1 phase must occur before cyclin E-CDK2 is activated late in G1 phase. So we then focused on the cyclin D and CDK4/6 since those are active early in G1. In panel D, you can see that, when we remove CCN2, cyclin D accumulates just fine. And Panel C shows that CDK4 is activated just fine as well. However, panel C also shows that the pRB protein does NOT get phosphorylated normally when we remove CCN2 from the cells. This told us that CDK4-cyclin D gets activated, but something stops it from doing its just job of phosphorylating pRB. And that is why cyclin E never accumulates, and CDK2 does not become activated.

To understand our next experiment, it is important to note that CDKs and cyclins typically do their work in the nucleus of our cells, not in the cytoplasm. They are often found in the cytoplasm when they are inactive, but when they become activated, they migrate to the nucleus and this is where most of their targets are. So, since removing CCN2 does not prevent cyclin D-CDK4 from becoming active, but it does prevent it from phosphorylating its target, pRB, one possibility is that cyclin D-CDK4 family to migrate to the nucleus when we remove CCN2. Indeed this is what we found:

At this point, we were now zeroing in on what is going on. When CCN2 is removed from cells, cyclin D still accumulates, and CDK4 still becomes active, but it does not migrate to the nucleus, so it does not find its target, pRB, and since that doesn’t happen, cyclin E never accumulates, CDN2 never becomes active, and the cells get stuck in mid-G1 phase. At this point, there was one obvious hypothesis – perhaps CCN2 is somehow involved in the migration of cyclin D-CDK4 to the nucleus of cells. If that were the case, we would expect that CCN2 must physically bind to cyclin D and CDK4 and we should be able to find them together inside the cytoplasm and the nucleus of these fibroblasts. Once again, our suspicions were correct:

As shown above, CCN2 physically co-locates with cyclin D-CDK4 as cells progress through G1 phase. And when this fails to happen, cells get stuck in G1 phase. The paper has one more figure, showing how the gene for CCN2 gets turned on by a transcription factor that is active in early G1 phase, but I think we’ve covered quite enough for one blog post.

To sum it all up, CCN2 is an important signaling molecule that gets made and released by collagen-producing dermal fibroblasts to help them multiply as they work to heal a wound. Our paper helped to reveal the mechanism of how CCN2 does this, by aiding cyclin D-CDK in migration to the nucleus, where it can find its major target, pRB, and usher the cells from early G1 phase into late G1 phase. The transition into late G1 phase is particularly important because this is when cells commit to completing a round of cell division. Once they hit a certain point in late G1, they cannot turn back because the process of DNA replication is now irreversibly poised to begin. Once that process starts, the cells must divide or die. This is because errors in DNA duplication can lead to cancer and so our cells have built-in mechanisms to protect us from that.

Congratulations if you made it all the way to the end of this little tour through one of my research articles! I hope you’ve enjoyed it, and I hope also that you can see why basic science research like this is so important. Only when we fully understand how the process of scar formation works can we begin to design possible interventions that would allow us to carefully control the process. We NEED this process in order to heal from any wounds or injuries that we sustain, but the same process also contributes to the scarring that causes some of our most stubborn diseases: COPD, heart failure, liver damage, kidney disease, and many others. If you want an end to these diseases, please support scientific research!