Originally published Oct. 9.
NEW YORK (GenomeWeb) – Long non-coding RNAs (lncRNA), molecules previously thought to play no role in cellular or disease development, may provide researchers a new avenue for developing a non-invasive diagnostic for prostate cancer.
Led by Ranjan Perera at Sanford-Burnham Medical Research Institute, researchers recently published a paper in The Journal of Molecular Diagnostics, describing their efforts to characterize the association between increased levels of lncRNA molecules and prostate cancer by genomically analyzing cell lines, tissues, and urine samples. They identified six lncRNAs that were overexpressed in prostate tumor samples compare to matched normal samples, "indicating the potential of using these lncRNAs to successfully distinguish cancerous from normal tissues," the study authors wrote in the paper.
First, Perera and his team looked to see whether lncRNAs were differentially expressed in prostate cancer cell lines versus normal prostatic epithelial cells using array-based tests. Out of thirty upregulated lncRNAS, subsequent quantitative real-type PCR analysis ranked three – XLOC_007697, LOC100287482, and SPRY4-IT1 – as being most significantly overexpressed. In particular, SPRY4-IT1 stood out, being upregulated more than 100-fold in prostate cancer cells lines compared to normal prostatic epithelial cells. This marker has also been found to be upregulated in melanoma.
Next, Perera's group compared levels of lncRNAs in a total of 28 prostate cancer tissue samples against matched normal tissues. In 10 pairs of fresh frozen tissues, microarray analysis revealed 10 up-regulated lncRNAs in prostate cancer tissues, while qPCR analysis of 18 pairs of formalin-fixed paraffin-embedded samples identified five lncRNAs that were expressed at higher levels in tumor tissues than in normal samples.
Based on these experiments, researchers homed in on six lncRNAs – SPRY4-IT1, XLOC_007697, LOC100287482, XLOC_005327, XLOC_008559, and XLOC_009911 – that were upregulated in prostate cancer patient samples and cell lines. Importantly, when the researchers tested urine samples collected from 13 prostate cancer patients and 14 healthy controls, they found the same six markers were also elevated. "The idea was to identify [biomarker] candidates … in patients' tissues compared to matched normal samples that are clearly differentially expressed," Perera, associate professor and scientific director of analytical genomics and bioinformatics at Sanford-Burnham’s Lake Nona campus in Orlando, told PGx Reporter.
In the paper, researchers noted that some normal matched tissues in their experiments had elevated lncRNA expression, which could be due to sample contamination. "An alternate explanation for this observation (elevated lncRNA expression in adjacent normal tissues) is that we observed the early transformation and spreading of prostate cancer to neighboring tissues that could not be detected by conventional histopathologic techniques," the authors wrote in the paper.
By investigating lncRNAs in cell lines, then tissues, and finally urine samples, Perrera's team gauged the feasibility of using these molecules to develop a non-invasive diagnostic marker in prostate cancer. "The rationale behind this study was to try to identify secreted lncRNAs in urine," Perera noted.
Previously, researchers had dismissed lncRNAs as having no functional importance. However, increasing understanding in the scientific community of key structural differences between lncRNAs and messenger RNAs (mRNAs), suggest otherwise. Studies have shown that there are four times more lncRNAs than coding RNAs. In studying these molecules, researchers have found that lncRNAS can impact gene transcription and expression, play a role in protein trafficking, as well as in post-transcriptional processing of mRNA.
In this latest study, Perera's group focused on lncRNAs, hypothesizing that these molecular markers would translate well to a diagnostic because they are more stable than coding RNAs, which "get destroyed in 40 seconds to a minute," Perera noted. "You won't be able to detect them."
If Perera and his colleagues are able to successfully translate this early research into a non-invasive lncRNA test, it will fill a need for more accurate diagnostics in prostate cancer. Rectal examinations and prostate cancer antigen (PSA) testing are commonly used screening methods in this disease setting. However, PSA testing hasn't proven to be a very precise tool in differentiating which patients are at risk for aggressive prostate cancer requiring surgery and which have indolent disease. Moreover, the National Cancer Institute estimates that only 25 percent of patients who are biopsied due to elevated PSA levels actually have prostate cancer.
Additionally, after an elevated PSA result, patients often require multiple biopsies, and in the end, most patients with a prostate cancer diagnoses have non-aggressive disease and could do well with surveillance. A non-invasive test, like the one Perera's group hope to develop, could becombined with PSA and other tests and help doctors avoid unnecessary biopsies for patients at low risk of aggressive cancer.
The next step for Perera's group is to study lncRNAs in urine samples from a large cohort of patients. Vipul Patel, medical director of the Global Robotics Institute at Florida Hospital and an author in the current study, has done close to 10,000 prostatectomies. Perera's team will try to collect patient urine samples from this cohort.
Companies such as Genomic Health, MDxHealth, and Myriad Genetics all market molecular diagnostics aimed at helping doctors garner more precise prognostic information on patients' prostate cancer aggressiveness. But all these tests analyze molecular markers from biopsy samples. Comparatively, a non-invasive urine test would be an advantage in the commercial setting by helping patients avoid unnecessary biopsies, but also useful in conducting larger validation studies.
Recognizing that clinical validation studies for an lncRNA diagnostic will need to enroll a significant number of patients, Perera pointed out that working with urine samples will make sample acquisition easier than collecting blood or tissue. "We are absolutely not worried about the number of samples," he said. "We can go up to millions [of samples]. Urine collection is not a problem at all." For example, Exact Sciences' clinical validation study for Cologuard, a stool DNA-based colorectal cancer screening test, involved 10,000 participants.
Perera ultimately envisions developing a next-generation sequencing-based lncRNA test using Illumina's MiSeq platform. No single lncRNA marker will work in this case, he said. "We have to come up with multiple markers and develop a multiplex test."
Perera's group will also study the functionality of the six lncRNAs with an eye toward therapeutic development. In the current study, using small interfering RNAs to knock down SPRY4IT1 in prostate cancer cells, investigators were able to hinder cancer cell proliferation and induce apoptosis. "That in itself tells us there is a function there and we can start to think about developing small molecules," Perera said.