Skip to main content

Biological News

Unstuck Proteins

This is a pretty interesting paper on several levels. It sheds light on Mucin I kidney disease (MKD), on protein degradation pathways (a hot topic these days, as those in the industry well know), and it also provides a small molecule lead compound. It’s a large multicenter team, starting off with the Broad Institute, but bringing in several other associated groups around Boston and Cambridge, as well as Wake Forest, Charles University (Prague), Pittsburgh, and the University of Cyprus. As you’d imagine from that, there’s a lot to be told, and I definitely won’t be covering every detail in this post.

MKD, a form of (fortunately rare) tubulointerstitial kidney disease, was already known to be localized to the MUC1 gene, via various frameshift mutations that end up in pretty much the same place in patients with the disease. The aberrant protein is missing part or all of important transmembrane and intracellular domains, so it’s no wonder it doesn’t work. It ends up secreted and presumably clogs up glomerular function somehow, but the details aren’t very clear (although as the disease progresses they can be seen histologically). There are no treatments, partly because the disease are quite rare and partly because of the lack of any mechanisms to get a handle on. From what I can see, most people don’t know that they even have MKD (or the other diseases in its class) until they start showing major kidney problems – and in fact, they may not even realize that they have a family history of it.

This new paper fills in the details, though. It is indeed a proteinopathy disease, although accumulation of the mutant Muc-1 protein takes a while to cause trouble. The unfolded protein response can handle it, up to a point (this is similar to the nucleolus stuff I blogged about last week – the cell has a lot of repair mechanisms, backups, and mitigation strategies, but they can only go so far). Eventually, slow loss of UPR function with aging and/or other environmental stress allows damage to build up. The team shows that a protein called TMED9 is a crucial part of the disease etiology – it’s a cargo receptor, part of a class of proteins that seem to sort proteins into forming vesicles inside the cell. And in this case, the mutant Muc-1 protein gets stuck in these TMED9-containing vesicles in between the Golgi and the endoplasmic reticulum, and they just pile up there.

The med-chem part of the paper comes in via a screen for small molecules that might disrupt this process, and the paper identifies BRD4780 as an active (shown at right). I’m not sure what I would have expected to come out of such a screen, but that isn’t it. That really looks like it should be hitting an ion channel, or maybe an amine GCPR. As a matter of fact, it was originally developed as a selective imidazoline receptor ligand, back in the 1990s at Allergan (where it was known as AGN-192403). The idea then was that such ligands could be antihypertensives, but to the Allergan teams’s surprise, this potent and selective I-1 compound seem to have no cardiovascular effects whatsoever. In years since, it’s turned up as having protective effects on mitochondria and as an inhibitor of fin regeneration in a zebrafish screen, but in neither of those was an imidazoline ligand expected to be a hit a priori. As in this paper as well!

Its mechanism of action on this protein trafficking problem is completely unknown. Something about it causes the mutant mucin to move on from the stuck vesicles, heading into endosomes and ultimately being degraded by the lysosome, which is just where such misfolded nonfunctional junk is supposed to end up. Update: as you can see from the comments, the paper tries to make a case for direct targeting of TMED9 by the compound, but I’m not really convinced by that, either. There’s no real SAR in this paper from what I can see, not even to the point of resolving the enantiomers of the compound, so there would seem to be some real room to step in and do something about some rare proteinopathies. There appear to be up to 20 different proteins that get stuck in a similar way on their way to the ER in various rare diseases, and the paper demonstrates that it works for two of these besides Muc-1.

The screen was a high-content cell-based one, using fluorescent markers to determine mucin protein localization. So discovery of this compound wasn’t waiting on the mechanistic details at all; this was a phenotypic screen that has both provided a compound and illuminated some of those details by doing so, which is just what a good screen is supposed to do for you. I hope to see follow up on this work, both out of scientific curiosity and in the hopes that this might lead to a useful therapy for a number of patients who currently have nothing available to them at all.

To that second point, though, one can also see how something like MKD has not attracted as much interest from the rare-disease-development crowd. One of the tricky parts of that business is that you have to be able to find your patients in the first place, and you also have to find them in time to be able to help them. A lot of potential targets drop off the list of prospects when you dig into those, and as mentioned above, this could be a tough one. People are born with the protein handling problem, but they don’t start showing symptoms for decades – and those symptoms are usually signs of rather serious kidney damage by that point. The cellular protein-clearing machinery (and the kidney as a whole) can take a lot of abuse, but that abuse occurs pretty silently, which means that the people who are likely to be helped the most by a new drug in this area don’t even know that they have a disease in the first place. Anyone wanting to do drug development in this field is going to have to find a way out of that dilemma. . .

24 comments on “Unstuck Proteins”

  1. Today is not the end of history. says:

    Perhaps, when we reach the day that all of us are routinely sequenced at birth, it will fix that last problem and spur efforts against these kinds of diseases. It sounds like the genetic frame shift is a pretty clear marker that someone should be on guard for MKD.

    1. loupgarous says:

      Private insurers and government health care providers stand to save considerable money by screening the population for even such rare diseases as this – if all such rare diseases can be screened for in a few blood draws, early in life. Early intervention is cheaper than late intervention if we’re talking about lifestyle changes rather than pharmaceuticals, and pharmaceuticals are cheaper than (in this case) kidney transplantation.

      1. Mike G. says:

        Sure, as long as the fix for the companies isn’t “dump this person from our plan (or don’t let them join the next plan) for a pre-existing condition”. My cynical tendencies make me think they’d see that as saving even MORE money.

        It was a plot point in the novel _FrameShift_ about Huntingdon’s, I think? The problem being that if the protagonist got tested, then he might not be insurable, but if he didn’t know, then the insurance companies had to take him. Is that what Huntingdon’s carriers really encounter these days?

        1. loupgarous says:

          One provision of PPACA that there’s broad bipartisan support for keeping is the requirement that pre-existing conditions be covered in health insurance plans. That’s been the case since 2014.

          1. cl says:

            No one will admit there against it, that doesnt mean they actually support it. The republican parties working very hard in the courts to get the entire set of legislation declared unconstitutional.

          2. loupgarous says:

            cl: large parts of PPACA are probably unconstitutional, things like taxes on gross revenue (as opposed to income) from medical device manufacture. The current Supreme Court is as balanced on that question as we’re likely to get, after decades of judicial activism, although those who whine about a shift to the right may actually have something to complain about if Ruth Bader Ginsburg’s health deteriorates badly before 2020.

            It doesn’t change the fact that there’s broad bipartisan agreement that pre-existing conditions shouldn’t be a disqualification from medical insurance coverage. Just as there’s bipartisan Congressional apathy about gargantuan price hikes on old medications because of grossly unforeseen consequences of laws meant to encourage safety and efficacy testing on orphan drugs. The hearings on that issue were brief and oddly without legislative effect.

          3. eub says:

            And 80%+ of identified Republicans support universal background checks for gun purchases. Heck, a slight minority of Republicans support legal pot. Yet I don’t see Mitch McConnell serving those Republicans somehow. It’s as if corporate megabucks affect political power brokers more than constituent will, sometimes.

            As I know you know, the economics are fundamentally problematic to protect people with preexisting conditions (and the ACA has done a real public service in highlighting the importance of this) simultaneous with allowing people to sign up for health insurance at any time, like after they come down with kidney failure. Republicans know citizens want to protect people with preexisting conditions, but they’re (by current consensus, I mean look at Romneycare for contrast) not prepared to do what it takes economically, so there’s a lot of hot air you ought not to fall for.

        2. Jonathan says:

          Good thing that’s explicitly illegal under GINA:

      2. nesprin says:

        Patients in end stage kidney disease are covered by medicare, so private insurers may have the opposite financial incentives…

      3. HFM says:

        For cost reduction, the time for genetic screening is before birth. There’s been a lot of success in preventing thalassemia in areas where it’s common (there’s a simple blood test for that) via discouraging carrier-carrier marriage and/or encouraging assisted reproduction with selection of healthy embryos. Genetic testing is cheap enough now that there’s interest in going after a wider range of diseases. To put it crudely, free testing for everyone, plus free IVF+PGD for every couple that shares a fatal recessive gene, is a lot cheaper than even a 25% chance of a doomed baby dying a sad, expensive death. The incentives are there.

        Insurance has also shown interest in pharmacogenomics, where the test might reduce the risk of adverse reactions to medication (which affects cost of care in the reasonably short term). But sequencing a baby just to find out about long-term risks…no. Even on the long shot that baby has something actionable, AND the recommendations are followed through with over the decades, baby will be an adult (and probably insured by someone else) by the time this starts to matter.

        Disclaimer: I work on a product that might be useful if you’d like to screen your whole country for a few hundred genetic diseases on the cheap. But there exists companies trying to make such products because there appears to be a market.

        1. metacelsus says:

          Since this disease in particular follows a dominant pattern of inheritance, discouraging carrier-carrier marriage isn’t relevant. Pre-implantation diagnosis would help, though.

    2. TruthOrTruth says:

      Great point, and Stephen Quake showed that we can sequence fetal genomes (i.e., before birth) so we need not wait that long! Link to article in my name.

  2. Anon says:

    The direct target claim is seriously weak sauce. They claim TMED9 as a direct molecular target of BRD4780 based on a cellular thermal shift assay, which could easily be observed by a number of indirect mechanisms. They also show a shift in the PAGE band migration, “consistent with possible covalent modification of TMED9 or other posttranslational modifications” but 1. They don’t do any proteomics to identify what the modified band is and 2. I’d be pretty surprised if that molecule is covalently modifying anything. In fact the band shift suggests that the thermal shift is probably _not_ indicative of direct engagement with the compound but rather indirect modification of the protein to some form with a different melt curve. I’m not sure how such a key claim is made with such weak support in a top-tier journal.

    1. Derek Lowe says:

      Yeah, I didn’t give any time to their mechanistic hypotheses, because I don’t find any of them convincing. The compound is certainly doing something that somehow involves the cargo protein’s function with mucin, but what, and how – mysteries.

      1. chemical biologist says:

        The problem is that their title focuses entirely on the mechanism: “Small Molecule Targets TMED9 and Promotes Lysosomal Degradation to Reverse Proteinopathy”. I also didn’t find any part of the mechanism convincing. Although the biology’s interesting, it’s disappointing that such chem bio claims made it into Cell, especially since that journal has had some past high-profile therapeutic examples that have later been disputed.

      2. Anon says:

        It’s not only that the evidence is too weak to support the claim but it seems like the path to the answer is relatively low hanging fruit. Identify the modifications in that band and other target classes (Deglycosylase? Dub?) present themselves. Maybe they tried and just aren’t reporting the results because it wasn’t a clean answer. If I had a research group in this area I’d already be generating that band for analysis, and I’d already have a draft paper started with the title, “XXXX, not TMED9, is the direct molecular target of BRD4780 mediating the clearance of pathogenic protein aggregates.”

    2. Lander says:

      Weak science goes to a top-tier journal all the time. It becomes more frequently nowadays.

      1. LondonLad says:

        Reviewers do your job. Properly. That’s All.

  3. Barry says:

    Shouldn’t MKD require BOTH mutation of MUC1 and dysfunction of the unfolded protein response?

    1. Derek Lowe says:

      I think that if you had inborn UPR problems *and* the Muc1 mutation, you’d be in very bad shape indeed. Infant death, perhaps? As it is, it seems that the disease gradually overwhelms the normal unfolded protein response; thus the late onset of symptoms.

      1. Barry says:

        I presume that w/o UPR, the embryo would not be viable. But if UPR is working, it should trigger apoptosis when overwhelmed, no?

  4. Cb says:

    Presumably, the simple basic compound BRD4780-like other alkalinizing compounds-nonselectively affects all acidic organelles including lysosomes, endosomes and the Golgi, leading to all sorts of biological effects includng the observation described.

    1. Anon says:

      Maybe, but they tested a number of compounds with basic amines and didn’t see this phenotype with any of the others. Could be this one has some feature that makes it localize to those regions more readily.

  5. Watson says:

    That looks amazingly like a sigma1 receptor ligand

Comments are closed.