My PI says I need to always keep in mind a narrative when writing-up a research manuscript.
So here goes:
The plan was simple and would make mincemeat of the problem overnight: Cure cancer. No, but in the ballpark of what friends and family anticipate at some point before thoughtful nods shift to metronomic ones. So, if not simple, then linear: develop a zebrafish model for t-cell leukemia. No quasi-evil N-ethyl-N-nitrosourea baths or bombarding gamma rays, just pure, organic, transgenic mutagenesis. To paraphrase Snapple, creating flavors Mother Nature never intended but should have. We had our not-fully-understood-but-we-know-it-works lymphocyte-specific, t-cell-signaling, stimulus-converting, protein-encoding gene with the tried and true capacity–and great homology–to express in primitive blood cell lines. With that as our construct backbone, we cloned in the bosses signature mutation-inducing insert. We cloned a few more of these constructs using different mutagenetic inserts for good measure…and why not? You see, injecting fish embryos at the one-cell stage with fragments of nucleotides that go by the names Stem-cell Leukemia (Scl), Leukemia-myloid oncogene 1 (Lmo1), and my personal favorite, Simian virus large-cell Tumor Antigen (TAg), set our fated fish on the longterm path to a tumor with fins. And what more significant way to give meaning to life than by predetermining not only its quality but course. And if I didn’t have my injected brood, population zero, then I’d still claim a clear Mendeallian half of their progeny. I’m able to think about these things and yet lose no sleep. Just to make sure I was dotting my “i”s and crossing my “t”s, I swapped a tumor gene for a jellyfish one and watched a seven day-old thymus light up like a uranium rod.
My ducks were in a row and everything was set, but then something unexpected happened. Well, at first, and for some time, nothing happened. Then something. Teleost thymus makes t-cells–the cell type in humans where a variety of leukemias form. Naturally, I would expect to see tumors blossom from my tiny fishes even tinier thymus. From our previous GFP experiments, we knew exactly where to look for the tumors because we knew the thymus. As we were waiting, we noticed one male after another started looking preggers. If I had instigated a mutation event in an autosomal sex-determining region in zebrafish then that would be something. Perhaps Sox9a, the gene in humans that inhibits the enzyme that converts androgens to estrogen. Clownfish, for example, are all born female. Through an as-yet-uncharacterized epigenetic event one will grow a set of testes, typically the smaller one in a pair. So there’s that. However, the general asymetry of these abdominal protrusions suggested tumor, not laden with eggs. As it turned out on closer, grosser, fatal examination, a sizeable tumor is what I did extract. Sometimes these tumor were so grotesquely out of proportion with the rest of their wetworks, they were the only identifiable mass in sight. Shoved to the corners and recesses were the intestines and liver. The thymus is nowhere near the abdomen. It resides up north between the gills, and posterior to a zebrafishes main congregation of ganglia–a few morphological leaps shy from an actual brain. Rationalizations and alternative explanations ensued.
Eureka! It must be the kidney, I thought. They’re in that region. Maybe the pre-kidney structure, the pronephros and its progenitor cells, proliferated out of control. Fish kidneys are above the intestines, between swimbladder and spinal cord. T-cell precursor cells originate in the kidney and later migrate to the thymus for further differentiation. In fact, endogenous Scl expression starts way back in the tail and winds-up in the kidney before turning-off.
Histologically, things were fuzzier. From sectioned tumor tissue slides we noticed unbroken cobblestone-like expanses of large blast cells (good) that didn’t look anything like mouse or human myloid cells (bad). Though there was this vascularizaiton that resembled a kidney in structure. More puzzling, however, was a preponderance of tiny cells with visibly little cytoplasm. When I say tiny, I mean they were pea stones next to the cobblestones, small enough to pass for a parasite, or maybe bacteria. My wild hypothesis, after scouring obscene pathology textbooks, was necrobodies: the corpses of fallen cells strewn about on my slide, the collateral damage from a tumor’s all-consuming single-mindedness. This notion lasted just as long as the time it took to convince the resident pathologist to put down his mouse slides and have a look at my fish.
Spermatids, not necrobodies?
But I know even less about spermatogenesis than I do about leukemogenesis. I’ve generated testicular cancer in these fish? It was like throwing darts at the bulls eye and hitting the ceiling. And so seminoma was the tentative name I arrived at for these tumors. I late found there to be a dirth of knowledge on the subject, but among those who discuss such things, the more general testicular germ cell tumor (TGCT) is the preferred nomenclature for what remains a largely uncharacterized zebrafish tumor. Some of the literature suggested that male zebrafish are particularly prone to developing TGCTs around two years of age. Out of sheer laziness I neglected to genotype most of the offspring from those parents who had my gene of interest integrated into their gonads. Only when a fish developed a tumor, did I screen it for the transgene of its parent. Fortuitously, this blinded me from selection bias against paying extra, special attention to the positive offspring and ignoring any naturally occurring tumors in the negative ones. I plotted a tumor incidence curve and there was a significant rise in tumors from the transgenic fish. I have to admit here that incidence curves where the rates is a percent on the y axis usually proceed in a downward curve. In other words, trenches not spikes. I stand by my decision to use the former because (a) it’s more visually stimulating (b) I define progress, generally an upward motion, as more tumors, not less.
I had a model for malignancy. However, zebrafish testicular tumors are a wide expanse of uncharted waters, made more difficult to navigate not only because I was the sole researcher in my lab studying zebrafish, but suddenly the only one not studying leukemias. While the zebrafish provides a superior transgenic model there are practical differences at the cellular level. For instance, teleosts, unlike mammals, have nucleated red blood cells and no hemoglobin. Basic molecular techniques were one thing–genetic homology is a humbling thing–but the histology of a cold-blooded, scaley, lungless, piscavore, a few billion years removed from humans, presents special challenges. For instance, fish cells are smaller and have a slightly lower specific gravity than human cells. This affects everything from flow cytometry to buffer solutions and transplants. Also, mouse researchers have their immuno-deficient stable mouse lines made-to-order. To immunocomprimise a fish, you need to either zap it (more specifically, it’s sperm) with radiation or soak larva in a gluccocordicoid. These are precarious measures, relying on toxic substances at minute doses for specific periods. Troubleshooting flawed experiments with mouse researchers proved at times difficult.
However, once the protocols were optimized–or worked most of the time–we sought to test the malignancy power of our designer tumors. When we spotted a pregnant-looking male swimming amongst friends and family, we often dissected him on the spot, harvested his bulge, mashed, filtered, counted and washed it, then backloaded it into a small-gauge insulin syringe. I raised a few wildtype fish for the express purpose of injecting them with these tumor cells. I aim for a major vascular region for maximum dispersal. A few of these cells, in a few of the fish, would hone to the testes and grow their own tumors, expressing my transgene all the while. This wasn’t enough, though. Most of the fish, most of the time should grow tumors from these cells. While the fish were indeed as inbred as English Royalty, apparently their innate immune systems detected enough foreign antibodies to kill-off most of the invaders. This would not do. As mentioned earlier, if I was a mouse researcher I’d be working with a model that had virtually no immune system to start with, and we wouldn’t be having this issue. Still, fish rule. Instead, I had to chemically ablate the thymus in my recipient fish. Unfortunately, the chemicals I was using don’t absorb well into adult fish. I needed to start with five day-old fish, soak them in dexamethasone for a few days, then inject their little bodies with adult tumor cells from a borosilicate needle I custom made for the occassion. Mortality was great, survivors few. But, with luck, a few grew up sporting tumors.
On these and their donors, we looked for genes common in human testicular tumors. There’s that homology again. We were interested in developmental genes, the ones promoting growth and regulating differentiation. These tissue-making workhorses are often highly conserved across the animal kingdom. Naturally at some stage in the animals growth and development, these genes need to be regulated and eventually halted. Sometimes, and the reasons are legion, they are brought back online or their regulators are taken out of commission, the results of which are what we recognize as tumors. That’s why it’s helpful to view cancer as an evolutionary disease.
Some of these genes have catchy names like Kit, ras, Oct-4, Sox9a, Nanog, vas, and some not so catchy names like MCFD2, and Wt1. If I can find a study that says Proml1, for instance, is a germ cell marker gene found in undifferentiated embryonic stem cells, specifically plasma membrane stem cell markers for neural and hematopoietic cells in humans, I have to compare its nucleotide sequence against the zebrafish genome for homologies sake. I like to see around 90% similarity. Sometimes, I’ll come across a gene charactertized to a specific cell type, as is the case with MCFD2 and sertoli cells. Great. Does this hold as true for fish as it does for humans? To find out we looked for expression of this gene in a host of zebrafish tissue: kidney, gill, intestine, liver, testes, and of course tumor tissue. Confoundingly, sox9a was expressed in everything except the gills and one out of three tumors. When I looked at the tumor under the microscope, it was one of the rare ones with a monolayer of spermatocytes and hardly any other cell type. Perhaps this is significant, since spermatogonia, like sertoli cells, are more primitive than spermatocytes and are known to express germ cell marker genes in humans. A tumor without these cells we could suppose might express MCFD2 very minimally if at all. Maybe, MCFD2 is a good genetic marker for spermatogonia in zebrafish. Still, its expression in other tissue is troublesome. We can’t rule out false positives, contamination of some sort or, since my samples were small, annoying randomness. And this has been the case with all the genetics markers: interesting but far from difinitive. Actually, this is the case for much of what’s published. Which is why I think I’m ready to finally write my manuscript!
PS. In 2002 hipsters co-opted the redneck trucker hat and wore it ironically. After that came mom jeans, granny glasses, track suits, Cosby sweaters, pregnant lady drop pants. What could possibly be next? What yet untouched mainstream style could ever outdo all that? Discuss
