The body of literature showing the therapeutic and diagnostic potential of microRNAs got a little bit bigger this week with the publication of two papers showing that the small, non-coding RNAs had a pharmacological effect when administered to non-human primates and that they be used to identify the tissue origin of various cancers, respectively.
In the first paper, which appears in this week’s Nature, investigators from Santaris Pharma demonstrated that intravenous administration of SPC3649, a locked nucleic acid-modified oligonucleotide, cut plasma cholesterol levels in monkeys. Additionally, the company reported no toxicities associated with the drug.
“This is the first demonstration that a microRNA can be targeted in a non-human primate species,” Sakari Kauppinen, Santaris’ associate director of microRNA research and senior author of the paper, told RNAi News this week. He noted that the company continues to expect it will begin a phase I, dose-escalating trial of SPC3649 — which is being developed as a hepatitis C therapy — in the first half of this year (see RNAi News, 2/21/2008).
In the second paper, investigators from Rosetta Genomics presented data demonstrating how a collection of 48 miRNA signatures could be used to pinpoint the origin of cancers of unknown primary with around 90 percent accuracy.
"Accurately identifying the origin of a metastasis in [cancer of unknown primary] patients can be critical for determining appropriate treatment,” Rosetta President and CEO Amir Avniel said in a statement. The research, which appears this week in the advance online version of Nature Biotechnology,“demonstrates the tremendous potential of microRNAs as effective biomarkers and is a significant step towards the development of the first microRNA-based diagnostic tests."
Rosetta is currently in the final stages of developing a diagnostic for identifying the origin of cancers of unknown primary, which is expected to be available in the US some time this year. The company also plans to launch this year two other miRNA-based diagnostics: one for differentiating lung adenocarcinoma from mesothelioma and one for differentiating non-squamous from squamous non-small cell lung cancer.
An early mover in the miRNA-targeting therapeutics, Santaris is banking on its LNA technology, for which the company holds the exclusive therapeutic rights, to give it an edge in the field.
LNAs are essentially nucleic acid analogs in which the ribose ring is locked by a methylene bridge connecting the 2'-O atom with the 4'-O atom — features designed to increase the molecules’ affinity and stability.
SPC3649 is an unformulated, single-stranded LNA oligo that targets miR-122, the most abundantly expressed miRNA in the liver and one that appears to play a role in cholesterol regulation, lipid metabolism, and HCV replication.
In the Nature paper, Santaris reported that “acute administration [SPC3649] by intravenous injections of 3 or 10 mg/kg … to African green monkeys resulted in uptake [of the drug] in the cytoplasm of primate hepatocytes and formation of stable heteroduplexes between the [LNA oligo] and miR-122.”
"Accurately identifying the origin of a metastasis in [cancer of unknown primary] patients can be critical for determining appropriate treatment.”
Furthermore, the researchers observed a depletion of mature miR-122 and a dose-dependent reduction in plasma cholesterol, which lasted seven weeks in the highest dose group. Cholesterol levels gradually returned to baseline over a period of three months following treatment.
Importantly, the Santaris team wrote in Nature that they “observed no acute or subchronic toxicities” in the SPC3649-treated monkeys aside from “transient, liver-biopsy-associated spikes in creatine phosphokinase, aspartate aminotransferase, alanine aminotransferase, and bilirubin.”
“To dissociate the effects of any liver-biopsy-associated toxicities in the safety evaluation of [SPC3649] treatment, two additional non-biopsy groups treated either with the high dose of [the drug or phosphate buffered saline], respectively, were included in the primate study,” the authors noted. “We did not observe any hepatotoxicity or renal toxicity in these animals as demonstrated by the absence of increases in the levels of the plasma transaminases ALT and AST, or of bilirubin, creatine phosphokinase, or creatinine after treatment.”
Part of the data from the Nature paper had previously been disclosed by Santaris officials (see RNAi News, 11/15/2007).
Although Santaris is initially developing SPC3649 as a hepatitis C treatment, Kauppinen said that the evaluation of the drug’s effects on cholesterol levels could provide an accurate indication of how the drug will work against the virus.
Additionally, he noted that included in the supplementary information accompanying the Nature paper is in vitro data from research conducted in collaboration with Stanford University’s Peter Sarnow showing SPC3649 to be “a very potent inhibitor of HCV replication.”
In 2005, Sarnow and his colleagues published a paper showing that sequestering miR-122 in liver cells triggered a “marked loss of autonomously replicating hepatitis C viral RNAs.”
In developing a miRNA-based diagnostic for tracking down the source of tumors, Rosetta developed a classification algorithm constructed as a branched binary tree, according to the Nature Biotechnology paper.
“In each node of the tree, classification proceeds to one of two possible branches, grouping together tissues with underlying similarities,” the paper’s authors wrote. “The decision at each node is a simple binary decision that can be performed using the expression levels of a few miRNAs. This scheme is analogous to a pathologist’s workup process, wherein a sample is assigned to increasingly finer subgroups through a series of differential diagnosis tests.”
“Basically, [the classifying system] follows a set of decisions where at each decision point only a very small number of microRNAs are relevant,” Ranit Aharonov, Rosetta’s executive vice president of research and development and head of computational biology at Rosetta who co-authored the Nature Biotechnology paper, told RNAi News this week.
“Most nodes of the tree are two to three microRNAs,” she said. “Basically, in each node, we look at the expression of these two or three microRNAs, check if their [expression] is high or low … then take either the right branch or the left branch.”
In the paper, the investigators measured miRNA expression levels in 400 paraffin-embedded and fresh-frozen samples from 22 different tumor tissues and metastases. Data from miRNA microarray analyses of 253 samples were used to construct the classification tree based on 48 miRNAs.
This classification approach had an overall accuracy of around 90 percent on blinded test samples and on more than 130 metastases, the researchers wrote. However, Aharonov noted that this figure is only an average that takes into account difficult-to-analyze tissues such as bladder.
“In most cancer classes we reached perfect accuracy,” she said, adding that the misclassification of bladder cancer was likely due to the small number of samples of bladder origin. “We’re working on many more samples … [which] we believe will enable us to optimize the algorithm and achieve better accuracy.”