Science1 distinct publisher2 min readPublished
The Hollings group puts lysyl oxidase inside the cancer cell, running its mitochondria, and reports that shutting it down leaves a backup survival pathway a second drug can close. The work is preclinical, and thinly quantified.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
Ozgur Sahin's group did not go looking for a compound that kills chemoresistant cells on the first pass. They went looking for a first move whose consequence is predictable: block LOX, watch the cells fall back on a backup survival pathway, then block that too [8]. Resistance in that framing becomes a scheduled step rather than the end of the road. Whether it holds depends on the adaptation being consistent enough to bet a second drug on, which is a stronger claim than showing that either agent does something on its own [2].
LOX has spent most of its scientific career on the outside of the cell, crosslinking collagen and stiffening the matrix around a tumor, which helps cancer invade and makes drugs harder to deliver [13]. The authors say it was poorly understood whether the same protein also regulates energy metabolism and mitochondrial homeostasis inside cancer cells [14], and they report the first evidence that it does [15]. Co-first author Burge Ulukan describes the effect as inhibiting several arms at once [17].
That prior history is also the analytical problem with an animal result. A LOX inhibitor that slows a tumor in a mouse could be working through the matrix [13] rather than through mitochondria, and the announcement of the combination result does not describe how the two routes were separated [6]. The paper's title commits to the intracellular mechanism [7]; the cell-level experiments that would license that commitment are where a careful reader goes first.
The quantitative content of the announcement is two adverbs. Growth was "significantly" blocked in "multiple" preclinical models, with no model count and no effect size [1]. The release also calls the approach a safe way to target chemoresistant tumors [18], set against chemotherapy's adverse effects and persistent tumor growth in the clinic [19]. Safe at this stage means only that nothing bad was seen at the doses used in the models tested, a far narrower claim than a toxicology package.
What the announcement leaves open is where the models came from: tumors that failed chemotherapy in a patient, or lines pushed into resistance in a dish. Those are different questions wearing the same word. TNBC is defined by lacking the three targets other breast cancers are treated through [9], and the authors name the subtype's metabolic heterogeneity and plasticity as the source of its aggressiveness and therapy resistance [10]. A strategy that treats that plasticity as something to route rather than outrun deserves attention, and for now it is a mouse-and-dish result [6].
Ranked by verification strength, evidence, and original report placement.
Triple-negative breast cancer may initially respond well to chemotherapy but often develops resistance to treatment.
The team discovered an unexpected role for the protein lysyl oxidase (LOX) inside TNBC cells.
Blocking LOX disrupted several processes on which the cancer depends, creating a weakness that can then be exploited using a second drug.
Experiments showed the combination strategy significantly blocked tumor growth in multiple preclinical models of TNBC, without chemotherapy.
Sahin: "It's a one-two-punch approach. First, we block LOX, which weakens the cancer cells. As they adapt and become dependent on a backup survival pathway, we deliver the second punch by blocking that pathway, too."
TNBC is named for cancer cells lacking three common targets used to treat other forms of breast cancer, leaving patients with fewer treatment options, with chemotherapy remaining a primary treatment despite significant side effects.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 31, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
science
Cells hoard polyamines to keep iron quiet, Whitehead team reports in Cell1 distinct publisher
product
Half of severe alopecia patients on upadacitinib hit 80% scalp coverage by week 241 distinct publisher
science
Salk's fly driver lines narrow octopamine access to single pairs of neurons1 distinct publisher
science
267 Ubiquitin Receptors, 93 Already Diseased: A Catalog That Gives Variants a Mechanism1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One publisher, one paper
Everything in this story arrives through a single pipe: GEN's write-up of the Hollings announcement, quoting only the lab. What keeps it from being thinner is a real destination — a Cell Reports Medicine paper with an exact title and named senior and first authors — so the mechanism claims can be checked by someone who goes and reads it. Nobody in this reporting has.
Bench, not bedside
The furthest this has traveled is mouse models and a journal page. There is no trial, no protocol, no patient. The single element with any real-world footing is leflunomide, already approved for another use, which lowers the cost of testing the pairing later and tells you nothing about uptake now.
"Safe" outruns the data
The word doing the most work here is "safe," applied to a combination whose entire tolerability record, as told, is no major weight loss and no signs of kidney or liver trouble in animals. "One-two punch" is a good metaphor; "significantly blocked" is not a result. The underlying biology may be exactly as interesting as claimed — the telling is simply a step ahead of what has been shown.
House compound, house announcement
The LOX-blocking half of the regimen is a compound the lab built, the senior author also runs science translation for Hollings' rapid-translation arm, and the news originates with the institution that owns both. None of that makes the finding wrong; it does mean the only voices in this story benefit from it being interesting. GEN discloses no patent or licensing position, so a reader can see the alignment but not measure it.
Sure who, unsure how much
Who did this, where it was published and what the proposed sequence is: settled. How big the effect is, across how many models, and whether it lasts: entirely open. That split is why the number lands mid-range rather than low — the attribution is unusually clean for a single-source story, and the magnitude is unavailable at any price from this reporting alone.