microRNA in Cancer
A look at the power of these tiny molecules in dogs with mammary cancer
After several Weekend Roundup posts in a row about “big picture” issues like the COVID-19 pandemic, the impact of the smartphones on our brains, and improving scientific communication, I’m going to take a beat to focus on a narrower topic: microRNA. This is also a chance to do something I haven’t done before and explain some of my own research.1
What ARE microRNA???
microRNA, as the name implies, are small molecules made of—you guessed it!—RNA. Researchers discovered that this class of nucleic acids had powerful control over embryo development in small invertebrates like the worm Caenorhabditis elegans in the 1990s. Soon, microRNA (sometimes called miRNA or miRs, produced “meers”) were discovered in virtually all species more complicated than bacteria, including plants, fungi, and all mammals. Adding to the intrigue, these molecules often had identical sequences of nucleotides (A, G, T, and C) in species as diverse as fruit flies, mice, and humans. Clearly, their functions were critical and highly-preserved.
What makes microRNA so unique is how it changes what we know about the “central dogma of biology.” This concept taught in high school biology holds that all creatures inherit DNA which determines the makeup of their cells and tissues. DNA in the cell nucleus is transcribed to an intermediate molecule, messenger RNA (mRNA), and this is used as a blueprint to make proteins, and these carry out of the vast majority of cell functions, from enzymes to the machinery to actually read and write DNA and RNA.
Unlike mRNA, microRNA does not code for anything, at least directly…Instead, they act as regulators of the whole process at multiple stages. miRNA fine-tune the expression of DNA genes (usually down-regulating expression, though not always) through multiple mechanisms. Sometimes they prevent translation by physically blocking the machinery that “reads” the mRNA (ribosomes). Other times, they directly bind to the nascent mRNA transcripts and cause them to be degraded.
What does all of this have to do with cancer?
While microRNA were initially discovered for their effects on the developing fetus, scientists soon found out that they had wide-ranging impacts on adult cells too, and could impact how much they grew and divided, how mobile they were, and whether or not they had specific functions. Not surprisingly, researchers in the early 2000s found microRNA expression was disturbed in most, if not all, forms of cancer.
One study found nearly 70% of cases of chronic lymphocytic leukemia (CLL) had down-regulated expression of two miRs. Then researchers found loss of different miRNA in colorectal cancers. After several of these narrow studies focused on just one specific type of cancer, a large study in Nature that looked at 20 different tumor types and many microRNA targets found the specific signatures of miRNA levels could actually differentiate the tumors.
These results got doctors and scientists wondering if microRNA could be used as a minimally invasive test for cancer. miRNA have multiple advantages over other types of testing. They are found in biofluids from blood to urine, cerebrospinal fluid, and more, so they would necessarily be less invasive than taking a chunk of tissue for a biopsy. Unlike some molecular targets like mRNA, microRNA are quite stable over time and at a range of temperatures and handling conditions. And in contrast to protein markers like Prostate Specific Antigen (PSA), they provide mechanistic insight and can be targeted by drugs. These concepts were confirmed by research in the late 2000s like this study that shows microRNA in blood could accurately diagnose prostate cancer.
Even cooler, microRNA are often found packaged within transport vesicles called exosomes. Research shows that these can be secreted by both normal and cancerous cells into blood and tissues to “communicate” with distant cells. Normal cells may use these to accomplish day-to-day tasks for metabolism and health, while tumor cells may use these to facilitate spread across the body.
Canine mammary cells shed exosomes that contain miRNA
The goal of my first PhD paper was to see if normal and/or cancerous mammary cells from dogs grown in culture produced exosomes, and if these contained microRNA. We were specifically interested in dogs and mammary cells for several reasons. First, there are numerous similarities in the appearance and biology of these tissues between dogs with mammary tumors and women with breast cancer. In theory, results from research in each world could benefit the other, a concept called “translational research” or One Health. Second, mammary tumors can be difficult to correctly diagnose on cytology (cells from some highly malignant tumors can look unremarkable, yet benign tumors can look ugly), meaning the current standard requires surgery for an invasive tissue biopsy. A minimally-invasive test for canine mammary tumors (CMT) would certainly be a game-changer.
We already had some evidence that CMT tissues contained abnormally expressed microRNA. And previous work in our lab showed that certain miRNA prevented normal tumor suppressor function in CMT cells. However, it was an open question whether these cells actually exported the microRNA outside the cell in a way that could be measured in blood or impact distant tissues. So first, we grew both normal canine mammary epithelial cells (CMEC) and CMT cells (from several variably malignant cells developed in our lab over the years) in test tubes. We took the cell culture media and removed large cells and debris with low speed centrifugation, then progressed to clearing out all but the smallest vesicles with ultracentrifugation (spun at 100,000 x the force of gravity!) When we looked at them with electron microscopy, we found abundant round, concave structures measuring 50-100 nm, in the right size range for exosomes. We check to see if these exosome-enriched media contained the protein CD9, which is a marker for exosomes, and they did.
Once we had identified the presence of structures consistent with exosomes we moved on to evaluate their miRNA cargo. We took cell-free fractions that contained these vesicles, extracted RNA, and performed deep-sequencing as well as quantitative PCR. Deep-sequencing revealed that both normal and malignant canine mammary exosomes contained hundreds of different types of microRNA. 145 miRNA were differentially expressed, and the signature could distinguish CMEC from CMT:
Many of these miRNA were predicted to regulate pathways involved in the cell cycle, apoptosis, exosome secretion, adaptation to hypoxia, and a shift towards more motility (important for metastasis). Several up-regulated microRNA, including miR-18a, miR-19a/b and miR-181a, were also predicted to target the estrogen receptor, and down-regulation of ER and other hormone receptors is a key event in the progression of malignancy in mammary tumors in both women and dogs.
microRNA in blood may help diagnose dogs with CMT
The cell culture experiments above established that it was at least possible for these cancer cells to be secreting exosomal miRNA. What we still didn’t know was whether this happened in real life, particularly whether these microRNA made it into the bloodstream. The goal of my second PhD paper was to take the concepts we identified in cell culture and see if they applied in live dogs with mammary cancer.
We recruited two groups of dogs for this study. One had biopsy-diagnosed mammary carcinoma. The other group was healthy controls that saw our Community Practice service (both spayed and ‘intact’ females were used for controls to account for hormonal variations). Then we drew blood and extracted RNA using kits optimized for small RNA. Similar to the first study, we performed deep-sequencing. This detected 452 individual microRNA, and 65 were differentially expressed between the cancer and control group. However, there was more overlap between groups, and the holistic signature was less able to differentiate cancer patients:
Nevertheless, select individual microRNA were much higher in the cancer group than controls. We evaluated a subset of the targets identified on deep-sequencing by absolute quantification with digital-droplet PCR (dPCR). The two most significantly up-regulated in the mammary carcinoma group were miR-19b and miR-125a. When we looked at their diagnostic accuracy by Receiver Operating Characteristic (ROC) plots, the Area Under the Curve (AUC) were 0.88 for miR-19b and 0.93 for miR-125a, indicating strong ability to differentiate the patients with cancer from those without.
An interesting side note: One dog in the healthy control group (subject HS3) was an outlier with extremely high miR-19b expression (32,364 copies/μL). Clinical follow-up on this dog revealed that within 1 year of this sample collection it developed widespread pulmonary metastasis from an unknown primary cancer and died shortly thereafter. The diagnostic sensitivity and specificity of miR-19b varied based on whether dog HS3 was included or not (we calculated it both ways due to the possibility this patient had an undetectable tumor at the time of sample collection and was incorrectly classified as “healthy”).
Finally, we wanted to see if any of these microRNA in blood were associated with specific tumor biopsy characteristics. It turned out miR-18a concentrations were significantly higher in dogs whose cancers showed evidence of lymphatic invasion (metastasis). miR-18a also trended higher in dogs with high-grade tumors compared to low-grade tumors, but this narrowly missed statistical significance. If these findings are verified by subsequent research, this suggests sampling microRNA in blood may not only assist in minimally invasive diagnosis of patients with CMT, but that it could also provide prognostic information about how advanced the cancer has become.
Putting it all together
Based on these studies and other published research, we came up with a proposed model for what effects blood miRNA have in dogs with canine mammary tumors. First, CMT secrete miRNA into nearby tissues and peripheral blood through exosomes and/or bound to proteins such as Argonaute (AGO). Then, these exosomal miRNA and/or AGO-miRNA enter peripheral blood, which could allow sampling for diagnostic use. Many of these miRNA, particularly miR-19b, can discriminate CMT patients from controls with good diagnostic accuracy. Next, computer modeling of the gene targets show that these exosomal and circulating miRNA (like miR-18a) likely regulate genes that alter hormone receptors (like the estrogen receptor ER), make tumor cells survive longer, increase proliferation, and promote metastasis and chemotherapy resistance. Some of these circulating microRNA correlate with clinical tumor behavior including lymphatic invasion (metastasis) and high-grade tumors. miR-18 and miR-19 genes are bolded in red to highlight their influence on all four processes and their proposed utility for diagnostic (miR-19b) or prognostic (miR-18a) purposes
These studies have a number of limitations. The first paper is based on in vitro cell culture work, and may not translate to live patients. The follow-up verified that a similar group of microRNA are dysregulated in blood as in cultured exosomes, but the groups were small, and the study was retrospective in nature and limited to single time-point sampling. Larger, prospective studies that track dogs with CMT and their miRNA levels over time and track their survival and response to therapy are necessary to verify the diagnostic and prognostic potential of these markers.
microRNA have been shown to be dysregulated in numerous other forms of cancer in dogs, including lymphoma, hemangiosarcoma, osteosarcoma, glioblastoma, and more. Anti-miRNA therapeutics are now in clinical trials in human medicine, with plans for veterinary patients as well. Much more research needs to be done to figure out all the mysteries of microRNA in cancer and how to use them clinically, but we are only in the beginning phases of our understanding, and it is an exciting time to be studying them!
On the off-chance you want an extremely deep-dive into the topic of microRNA, exosomes, and translational mammary tumor biology (or possibly need help treating recalcitrant insomnia), you can read my full dissertation here. It includes the core studies discussed in this post, as well as a detailed review of the basic science and medical (human and veterinary) literature and some research we did not publish.
Awesome and I understood only 5% of it!