Skip to main content

A Plan For New Antibiotics

The Pew Trust has come out with a document that I hope makes an impression: their Roadmap for Antibiotic Discovery. Every year the the bacterial resistance problem gets a bit worse, and the most we can do is make it get worse at a slightly slower rate. As long as we want to kill bacteria, we’re going to be in an arms race with them. (I’ve actually done antibiotic drug discovery myself, and killing bacteria (without killing everything else) is very hard indeed).

You can find quite a few people who are convinced that Big Pharma, unable to make their customary billions, have callously abandoned every one to bacterial death, but as I’ve written several times before, it didn’t quite work like that. Several large companies did pull out of antibiotic research, for sure, but several didn’t. The companies that didn’t pull out, though, never made much progress, and it’s the combination of poor prospects and return on investment that has dragged the field down. If you went to a big drug company and asked if they’d like to have a new drug like linezolid or ciprofloxacin, with the prospects that each had when they first hit the market, they’d sign right up. But divide that by the chances of getting one of those, and it’s not a good idea. Put simply, if the antibiotics divisions of GSK or AstraZeneca (to pick two that kept going) had been separate independent discovery companies, they would have been the largest things in the field. And they would have gone bankrupt. Years of work in both shops led to no marketed drugs – the money flowed out constantly and nothing ever flowed back in.

The Pew roadmap sums up the situation this way:

New discoveries dropped precipitously from the 1980s onward. As a result, the development of antibiotics has declined, with new Food and Drug Administration (FDA) approvals for these drugs falling from 29 during the 1980s to nine in the first decade of the 2000s. All antibiotics approved for use in patients today are derived from a limited number of types, or classes, of antibiotics that were discovered by the mid-1980s (Figure 1). This is even more concerning than the decline of drug approvals because resistance to one antibiotic often leads to resistance to multiple antibiotics within the same class. While drugs can be categorized or classified in a variety of ways, for the purposes of this document, antibiotic classes are based on similarities in chemical structure.

Faced with poor discovery prospects and diminishing returns on investment, major drug companies have cut back or pulled out of antibiotic research altogether. This has left much of the remaining discovery work to small, “pre-revenue” companies with no products on the market and limited budgets and R&D capacity. Most industry antibiotic development programs are primarily focused on modifying existing classes of drugs discovered decades ago to circumvent bacterial resistance and better target difficult-to-treat infections. Though essential, such incremental advances are not likely to meet the looming public health challenge of antibiotic resistance in the long term.

This is true. And one of the things I like about the Pew Roadmap is that it recognizes that the biggest barriers to finding new classes of antibiotics aren’t really regulatory or financial ones: they’re scientific. The plan calls for efforts to produce new chemical matter in the property space that antibiotics tend to occupy, which is a rather weird one compared to most other drugs and not well served by current screening libraries:

It is important to clarify that this proposed effort is not focused on broad expansion of synthetic libraries or seeking out new sources of natural products. Instead, the goal is to execute focused work to carefully vet existing chemical matter and conduct targeted synthesis and modification of new chemical matter based on what is known about antibacterials from published and unpublished sources, incorporating insights and guidance as new research findings emerge.

It will take time to determine whether conditional guidelines can be developed for Gram-negative drug entry and efflux avoidance. Meanwhile, practical and transparent knowledge-sharing mechanisms should be established to better inform discovery scientists on how to identify new chemical matter based on drug-like qualities and what is already known about the chemical properties of antibiotics. This would require collaborative research to facilitate new and directed approaches to generate and modify chemical matter. For example, working together across multiple disciplines, scientists may begin to develop new semi-synthetic antibiotic templates derived from fragment-based or natural products-based starting points that better target Gram-negative or Gram- positive bacteria.

There’s also a proposal to revisit failed natural product projects that showed some antibacterial promise, in an effort to see if the reasons for these failures can be overcome with semisynthetic derivatives and so on. The roadmap also calls for a concerted effort to work on targeting Gram-negative organisms in particular, and to do that we have to have a better understanding of their membranes, transporters, and efflux pumps. These are the barriers that make those organisms so difficult, and it’s clear that we don’t have a good enough handle on them. (For example, there are two layers of membranes around these creatures, each with different properties, so it’s hard to come up with a compound that will get across both of them at once. And each of them are lined with membrane-spanning proteins that pump a huge variety of structures – pretty much anything that’s not on the guest list –  back out of the cells).

The team calls for a joint academic/industrial effort to (1) first identify what we actually know about targeting these organisms, with opening of the books on past failures, (2) come up with standardized assays to measure drug penetration and measure existing compound collections against them, while (3) at the same time working on efflux pump and membrane disruption ideas. To that first point:

There is growing concern that as industry teams are downsized or shuttered, antibiotic scientists have moved to other firms, shifted to different biomedical areas, or retired, leading to the loss of valuable institutional knowledge and expertise. Antibiotic discovery has a long history, but much of the published research is buried in old journal issues or out-of-print books, and other research never makes it to publication. Organizing this body of research and making it accessible to the scientists who need it is critical for advancing discovery. Valuable knowledge may include compilations of screens that have been run before and information on past research programs. While much of this information is publicly available, what may be most useful is an account of what projects failed, and why.

Outside of these traditional antibiotic ideas, the document also has a section for working on some of the less standard ones: going after virulence factors, for example, or using immunotherapy approaches, drug delivery modes that haven’t been explored enough in antibiotics, predatory bacteria (an area that’s beginning to be explored with DARPA funding), etc. There are a lot of areas that have seen work on and off (the polymyxin membrane-disrupting compounds, iron-uptake pathways as a route to drug entry), that need a more concerted effort. For many of these, we’re going to have to figure out what appropriate development pathways will even look like.

Another area that gets attention is something that you’d think would have been explored more in antibacterials, but hasn’t: combination therapies. This is the rule in antivirals – you hit several mechanisms at once to lower the chance of resistance, but you really don’t see mixed-mode therapies in antibiotic work as much. The roadmap calls for an exploration of this area, with special attention to trying to revive the single-target agents that have failed over the years for lack of efficacy on their own.

There are many more details in the plan itself, so if you’re into this sort of thing I encourage you to read it. The whole thing seems very sensible and well-aimed, but the tough part now will be implementing it, because it won’t be easy and it won’t be cheap. One encouraging thing, though, as a correspondent put it, is that when it comes to Congress, “there’s no constituency for multidrug resistant organisms”. If some prominent politician wants to make a Joe Biden-like splash in addressing a major public health problem, this looks like an excellent place to do it.

112 comments on “A Plan For New Antibiotics”

  1. Barry says:

    It’s too easy to say “the biggest barriers to finding new classes of antibiotics aren’t really regulatory or financial ones: they’re scientific”. The scientific barriers are real, and impressive, but surely they’re less than were overcome to develop e.g. the first HIV protease inhibitor drugs at a time when we knew almost nothing about anti-virals.
    The financial barrier to discovering and developing new antibiotics may not be fixable within the current business model. If you were to present our FDA a novel antibiotic today that is effective against all multi-drug resistant organisms currently known, the FDA–for excellent reasons–would want to approve it only for the narrow indication of third-line treatment of multi-drug resistant infections. In this way, it would remain useful longest. Alas, that can’t work. Long before the patent protection had run out a pirate company (historic precedent suggests in India) would be producing it, and selling it, and–inevitably, it seems–a resistant infection would defeat it long before you made back your costs of R&D.
    The existing drug company paradigm in which patent protection grants the innovator a period of exclusivity to recompense and reward the risky investment in drug discovery and development has sufficed to build the business we know today. But the next generation of antibiotics may require a whole new funding concept.

    1. Derek Lowe says:

      I think you’ve got a good point about the business/IP models. But as for the scientific barriers, it’s true that the early HIV stuff went in when we knew comparatively little about antivirals. The problem is that we know comparatively a lot about antibacterials, which has only served to illustrate just how hard they are.

    2. ScientistSailor says:

      The outer membrane of Gram(–) bacteria make this problem orders of magnitude harder than anti-virals…

      I was very happy to see the emphasis on combinations in the document, this must be a high priority.

    3. Dan C says:

      The value/profitability of new antibacterial drugs must also be tied to the severity of the drug resistance problem. If you can’t make your money back on your novel broad spectrum antibiotic because the FDA will only approve it for use in rare instances, this seems to imply that there is not a substantial unmet medical need. If the antibacterial drug resistance is now (or soon will be) as severe a problem as experts are saying, then the financial incentives for developing new antibacterial drugs should be growing. When untreatable MDR infection become common enough, people will be willing to pay top dollar for effective new antibacterial drugs, and the necessary innovation will occur.

    4. tangent says:

      Your point about narrow indication versus OTC piracy is damning and depressing. That type of flat-out abuse ought to be totally unacceptable, decouple it from intellectual property yada yada, this is essentially a biological WMD threat.

      What if insurers were willing to pay a million dollars for a lifesaving course of third-line antibiotics, would that make things pencil out? Not that I *like* that scenario, but it would make more sense than paying hundreds of thousands for cancer drugs that give three months.

  2. Google gave me nothing says:

    Tangential question: What’s the half life of antibiotic resistance?

    Say you have a resistant population, if it doesn’t come into contact with the drug then there’s no selection pressure on maintaining those genes as are. If they mutate without penalty for long enough, it’d cause loss-of-function. How long would that be? Would that mean that shelved antibiotics could eventually have a resurgence? Or is the timescale so ridiculously long that the cockroaches could evolve to be the dominant intelligent lifeform on Earth and have to find antibiotics of their own?

    1. Me says:

      Ur too smart for this thread :O

    2. Barry says:

      turns out, the fitness cost of drug resistance varies from mechanism to mechanism. For some mechanisms, the adaptation washes out of the population when exposure to the drug is removed. For others, it persists:

      “several drug classes and species of bacteria on average did not show a cost”

      ” Clinical studies have shown that in some cases, resistant bacteria remained abundant in the population (Enne et al. 2001; Sundqvist et al. 2010) or even increased in frequency (Arason et al. 2002) despite the absence of drug, while in others the proportion of resistant bacteria within the population declined (Seppala et al. 1997; Austin et al. 1999; Bergman et al. 2004; Gottesman et al. 2009), as expected. In epidemiological studies, reducing the use of antibiotics often leads to a reduction in the frequency of resistant strains, but it rarely succeeds in eliminating them altogether (Salyers and Amabile-Cuevas 1997; Andersson 2003; Enne 2010; Johnsen et al. 2011).”
      “most resistance mutations in bacteria confer a fitness cost. This result is not surprising as many antibiotics target important cellular processes and resistance to them either disrupts those processes or imposes large energetic burdens that reduce competitive ability against sensitive strains. However, we have also found that there is substantial variation in fitness costs among species and drugs. This variation is large enough to include, occasionally, what might be classified as no-cost resistance mutations”

      1. Google gave me nothing says:

        Thanks for the reference! Agreed, the half-life would depend very much on the metabolic/fitness cost of expressing resistance genes, therefore it’d be species and/or method of resistance specific. I’ll get reading that paper 🙂

    3. Kriggy says:

      No idea but I suppose that could be interesting reserach but I dont think that the resistance gets away fast enough.

    4. This doesn’t answer your question, but you may like the paper as it is related. Your comment reminded me of the following manuscript (see the yin and yang symbol associated with publication 113, link in my name

      Deoxynybomycins inhibit mutant DNA gyrase and rescue mice infected with fluoroquinolone-resistant bacteria
      Parkinson et al. Nat. Comm 2015

  3. Me says:


    I started out as an antibac medicinal chemist in a big pharma department. Most of the guys with the knowledge in that dept have long-since retired and I’m moved on from chemistry in general.

    Glad to see they’re not pushing modern target-based design as a strategy!

    I reckon they have a point on clinical paradigms, and the immunological side of it is huge. Although my second med chem job ended as a result of the dismal failure of a therapy targeting immunological side-effects of infection.

    None of this is easy – and agreed the system is not currently set up to incentivize antibac work either.

  4. exGlaxoid says:

    I don’t think that the barrier to discovery would be that large if there was a clear payoff for the discover. If someone said that they would pay $1 billion to any group that discovered a new antibiotic with low resistance, that meet certain criteria, it might convince a few companies to at least run some simple HTS’s and I would would bet that we could find ten to a dozen new candidates within a few years. Antibiotics have nice, phenotypic assays, that are not perfect, but are much simpler to run than for lupus, RA, MS, or most other diseases, there are many accepted models in animals, and there are simple end points.

    I suspect that almost every company has at least one or two antibiotics that they found and tabled due to economics, if the NIH, FDA, Pew, or a third party offered to buy those assets and test them further, many companies might be willing to sell or license them for that type of work. But I don’t think it is likely to happen, because no president is going to declare a “moonshot” on antibiotics until there is a serious need for it. And Barry is absolutely right, that even if we try to keep a new drug for last resort, someone will be making it within a week elsewhere, so the key is to discover a new series every year or two and just license it to the world cheap and keep a project going forever.

    1. Derek Lowe says:

      The animal models are great, and the phenotypic assays are pretty good too, as you say. But I’ve seen some large-scale screening efforts come to zilch in this area – I really don’t think that just running some HTS screens will do the trick.

      Gotta disagree with you on the tabled antibiotics, too. Why not outlicense these things and pick up some money? What I do think are sitting in the files are a bunch of compounds that got up to a certain point before running into problems, and those, I think, are what the Pew people would like to see unveiled more.

      1. Vaudaux says:

        “…sitting in the files are a bunch of compounds that got up to a certain point before running into problems, and those, I think, are what the Pew people would like to see unveiled more.”

        I agree with Derek. If anyone knows of antibiotics ready to go into human trials, please speak up.

        A problem with unveiling the tabled scaffolds is that information about them is hard to come by. Even if the company still exists, data will be scattered across dozens of handwritten notebooks, or on spreadsheets saved onto floppy discs. The scientists who could tell you why a project was dropped or why it should be resurrected are also scattered – across the industry, many now semi-retired, some now consulting. One part of the Pew roadmap is collecting old information before it is completely lost.

        1. lynn says:

          Vaudaux is right – the information on tabled programs will be hard, or even impossible, to gather – yet there is a push to do so. As to the antibiotics sitting on the shelf – perhaps there are a few – and most of those almost-clinic-ready have been outlicensed to small pharma. Big Pharma was in this business for many years, employing hundreds, thousands of scientists in this particular field. The efforts – and they were major – failed to turn up useful novel classes of agents. So, yes, what was developed were modifications of the old classes, especially to overcome resistance. And that’s often used as an indictment of Pharma – that they just made me-toos. That was what worked. Novel developable antibacterials are hard to come by. Also – the whole movement to genomically defining new targets [which most companies moved to] was a bust…for a few reasons. Largely because we knew the essential genes already (pretty much) via microbial genetics, but not how to get inhibitors into cells (Gram(-) especially) and not how to avoid resistance. This is discussed in the Roadmap which gives references.

    2. ScientistSailor says:

      GSK and AZ have published retrospectives on their antibiotic screening effort. Close to 150 HTS campaigns between them resulted in NO clinical candidates. More HTS is not the answer…

      1. HTSguy says:

        Do you happen to have the references for those?

        1. Vaudaux says:

          – Payne, David J., Michael N. Gwynn, David J. Holmes, and David L. Pompliano. “Drugs for Bad Bugs: Confronting the Challenges of Antibacterial Discovery.” Nature Reviews. Drug Discovery 6, no. 1 (January 2007): 29–40. doi:10.1038/nrd2201.
          – Tommasi, Ruben, Dean G. Brown, Grant K. Walkup, John I. Manchester, and Alita A. Miller. “ESKAPEing the Labyrinth of Antibacterial Discovery.” Nature Reviews Drug Discovery 14, no. 8 (August 2015): 529–42. doi:10.1038/nrd4572.

          1. HTSguy says:

            Thanks! Interesting and a bit depressing.

  5. Dr. Manhattan says:

    ” it might convince a few companies to at least run some simple HTS’s and I would would bet that we could find ten to a dozen new candidates within a few years.”

    Been there, did that. Glaxo has published extensively on the failure of multiple (like 75-80) HTS programs to identify any new compounds that could be advanced. AstraZeneca published a similar result; and that has been the experience of several other companies as well. HTS chemical libraries in big pharma are not the place to search for new antibacterials. Discussing that problem in detail is a very big part of the Pew document.

  6. Markus says:

    I hope that CRISPR will be a game changer to fight antibiotic resistance.

    1. Barry says:

      and buckyballs!

      1. PS says:

        … and stuff prefixed with “nano”.

    2. Hearsay says:

      Not too long ago I heard Stuart Schreiber (jokingly) say something along the lines of “CRISPR is the last refuge of the incompetent”

  7. John Wayne says:

    I heard the phrase ‘CRISPR everything’ recently, made me laugh.

  8. Dr. Manhattan says:

    “I hope that CRISPR will be a game changer to fight antibiotic resistance.” Sure, no problem! All you have to do is deliver RNA & an enzyme into intact bacterial cells. Hey, maybe Moderna can help…

  9. bacillus says:

    If antibiotics simply didn’t exist chances are we’d have been forced to develop vaccines against many of the most problematic bugs. Sure, some academic groups have focused on e.g. group A Strep, Staph and Klebsiella vaccines, but these too would need $B to get them to market. Absent antibiotics, we might have already made appropriate headway with vaccines against some of them. Unfortunately, antibiotics proved to be so effective against so many pathogens for so long, that the vaccine path wasn’t followed for most of them.

    1. Barry says:

      vaccines (along with potable water and sanitary sewers) are among the only bargains in all of healthcare; everything else is vastly more expensive. They are not, however the answer to everything. A hundred years of trying has not produced e.g. a tuberculosis vaccine that works. That doesn’t mean it’s impossible–there is recently progress towards vaccines against falciparum malaria–but it’s unlikely that vaccine will obviate the need for anti-infective med. chem.
      Another avenue that we have neglected since the advent of antibiotics is therapeutic ‘phage. It doesn’t fit neatly into our current regulatory paradigm. There is no well-defined therapeutic agent, and–because it mutates constantly–it’s hard to know that last year’s Clinical results predict anything for the dose given this year.

      1. tangent says:

        Could we in principle manufacture virus from a fixed genome, if serial mutation of the strain is a blocking issue? Generate accurate genetic material and build into an infectious DNA clone? Only a few generations worth of serial replication should ever be needed.

  10. Anon says:

    If this is really important to society, then we just have to align the incentives so that it is also important to pharma and biotech:

    1. Provide more government funding for basic research in this field
    2. Increase patent lifetimes for new antibiotics
    3. Provide a minimum revenue guarantee before the need arises
    4. Encourage more antibiotic combinations and disallow use of single antibiotics to reduce gradual build up of resistance
    5. Etc.

    If we are not prepared to do this, or if the political and financial willpower is not there, then by definition, it is not an important issue to society. It’s not rocket science.

    1. I’m not sure I can agree with this interpretation. It comes back to the issue one sees in economics in which “Homo economicus” is the perfect consumer, always weighing all available data, always doing things in her best interest. This isn’t reality. In reality, we have a sizable number of people wanting to vote for a certain candidate even when the people don’t agree with many (or all) of that candidate’s public (and chameleonic) positions, just because of their “gut feeling” that that person is a straight talker and will really show ’em!
      You could make the same argument around global warming. Clearly this is existentially important to humans, but at this point too little is being done.

      I do agree that aligning incentives would help, but I don’t know that lack of alignment suggests lack of importance.

      1. Anon says:

        It was meant as a provocative shock tactic, i.e., “prove your claim that it’s important by putting your money where your mouth is”.

    2. Andy II says:

      There are already incentives.

      1. Provide more government funding for basic research in this field
      – NIAID, BARDA, DARPA. GSK and several biotechs obtained huge fundings for their programs. As often the cases with those government funding, they did not produce good return yet.

      2. Increase patent lifetimes for new antibiotics
      – Generating Antibiotics Incentives Now (GAIN) Act is effective to provide pharma companies who are developing anti-infectives for prioritied indications. It gives first track status and extra 5 yr exclusivity. It means that no IP protetion for the drug itself in the US, if it is an NCE, it will get 10 yr exclusivity in US (5 yr for NCE and 5 yr for GAIN). A list of drugs designated as a Qualified Infectious Disease Product (QIDP) under the GAIN Act is available at PEW foundation.

      3. Provide a minimum revenue guarantee before the need arises
      – GAIN Act is meant to be.

      4. Encourage more antibiotic combinations and disallow use of single antibiotics to reduce gradual build up of resistance
      – One of the problems with this approach is if someone spent money to come up with a novel combination with “existing” antibiotics, doctors/pharmacy will see the publication and use their pharmacy stocked drugs. No need to pay the innovator, technically. Even if we know some deficiencies with “existing drugs, we do not want to improve them because there is not much difference with the old one and the improved one as doctors are already familiar with the old ones.

      We talk about resistance. And if you check the epi data of strains from clinical samples, you notices about 30% are resistant to the current first-line drugs. But think about this. Even state-of-the art cancer immunotherapy can treat only small population of patients.
      There were already breakthrough in antiinfectives and we are waiting for next one to come, which needs time and money and patient (not who suffers from infections). Let’s put the stupid excuse of “best for the shareholders value” second.

  11. mardukofbabylon says:

    Are Viruses targeting bacteria an option ?

    1. Vaudaux says:

      See above, Barry in reply to bacillus at 12:55 pm. Viruses targeting bacteria are bacteriophage, or ‘phage. Lots of work in Russia with anecdotal reports of success but no controlled clinical trials.

  12. Anonymous Researcher snaw says:

    Back in the 1990s my former employers, like many others, made a substantial investment in using microbial genomics to identify new drug targets. Without going into details that would identify the company, suffice it to say I applaud the Pew Trust for its ideas because I saw many of the challenges they describe from the inside. Anybody with an interest in the area of antimicrobial drug discovery should “read, mark, learn, and inwardly digest” the Pew Trust’s report.

    I am listed as an Inventor on a patent related to antifungal research; in some ways that’s an even tougher challenge because fungi are eukaryotes and thus have greater biochemical similarity to mammalian cells than do bacteria. Most antibacterials bind to something not found in eukaryotic cells.

  13. Glen says:

    Suggestion: Income from successful new antibiotics will be tax free as long as resistance remains low for the targeted bacteria. This would provide financial incentive for both discovery, and stewardship of use.

    1. Lyle Langley says:

      @ Glen,
      Unfortunately that does not provide an incentive for stewardship for overuse. The doctors don’t benefit from the tax free revenue. Potentially the owner of the drug could try and limit the use, but that’s very difficult to do – it doesn’t happen now.

      1. loupgarous says:

        Antibiotic stewardship’s an issue properly addressed by prescribers and users. Farmers could be responsible for much proliferation of antibiotic resistance plasmids, from their use of antibiotics in feed as a means of getting good weight on animals without free-ranging them.

        One major US chicken processor has a disingenuous series of television commercials “debunking” the “antibiotic-free” label on organically-grown chicken, leaning on the slender reed that “by Federal law, all chicken must be ‘free of antibiotic’ before going to market”. They don’t say ‘free of antibiotic resistance plasmids,’ a whole other matter. Not that their chicken’s better or worse than average in this way, just that they’re playing hide-and-seek with the facts on television. I’ll have my chicken and eggs well-cooked, thanks.

  14. Curious Wavefunction says:

    “When it comes to Congress, “there’s no constituency for multidrug resistant organisms”.

    But there’s lots for multilogic resistant human beings.

    1. loupgarous says:

      John Oliver’s expose on the way big chicken packers (sounds obscene, doesn’t it?) abuse the farmers who actually grow the chicken Americans eat goes into great detail describing one of the key constituencies for multidrug-resistant organisms – those people in Congress who, irrespective of party, have had the big chicken processors’ back. His name for these legislative stalwarts is a portmanteau of “chicken” and the plural of a noun beginning with the letter “f” indicating one who engages in coitus.

  15. kjk says:

    Could we make yeast compete with bacteria, forcing the yeast to evolve antibiotics? What kinds of environment maximize the evolution-rate? As long it is is faster than the rate of evolution in the “wild” (hospitals, etc), we may be able to crank out new antis more quickly than resistance appears. It will also help to slow down the evolution rate in the wild by improving hospital hygiene.

  16. lynn says:

    Bacterial viruses [bacteriophages] have been looked at as therapeutics for almost 100 years. The Georgians have used them, mostly for wound infections, for many years, but there have been few controlled trials. More recently this avenue has been revisited. But – as with any of the approaches people have come up with, there are many obstacles to the practical use of phage therapy for systemic infection. This and other alternative therapies are mentioned in the Pew Roadmap. What the roadmap proposes is that, for all the alternative approaches, what is needed is guidance on how to proceed in developing such therapies or adjuncts. What experiments to do? What problems have been encountered? Another problem that the roadmap addresses is that, while there is a lot of literature out there on antibacterial discovery, most newcomers to the field seem to think nothing has been done. So, a way of bringing together that knowledge base, literature, papers, books and making them available [or at least pointing to them] so that people don’t keep on reinventing the wheel. Yes, of course, things my be done in better ways with new tools, but don’t think, as Dr. Manhattan notes, that most chemical libraries haven’t been screened [by all sorts of methods]. It’s easy to kill bacteria – but not without killing the patient.

  17. watcher says:

    Have to wonder if going into patents (and even allowing infringing) would provide the most useful “leads” that could be obtained.

  18. Mark Thorson says:

    Never fear! Watson will save us!

  19. DCRogers says:

    “the biggest barriers to finding new classes of antibiotics aren’t really regulatory or financial ones: they’re scientific”

    led me to

    “the biggest barriers to keeping old classes of antibiotics aren’t really scientific; they’re regulatory or financial”

    Basically, overuse encouraged by regulatory or financial incentives.

    For example, you cannot use in animals antibiotics that are used in humans (sounds good!) but you can use compounds that are so closely related that any resistance gene that develops to the animal antibiotic might already provide the kiss of death to its human counterpart, or at least an awfully good head-start towards one.

  20. gippgig says:

    Very interesting article:
    Assembly and clustering of natural antibiotics guides target identification
    Nature Chemical Biology Vol. 12 p. 233 doi: 10.1038/nchembio.2018

  21. Santa Maria says:

    What about immunotherapy targeting particularly nasty strains? Could the gram neg double membrane serve as a target?

    1. Barry says:

      the lipopolysaccaride (LOS) that comprises Gram-negative outer membrane is indeed immunogenic. It is believed that cross-reactivity of antibodies to LOS are responsible for some human auto-immune disorders.

    2. J Santa Maria says:

      I can’t help my curiosity of who else could be a Santa Maria that follows this blog and is actively interested/involved in antibiotic research…

  22. Barry says:

    Without any numerical analysis, it jumps out that more than other therapeutic areas, lots of antibiotics violate Lipinski’s rules. That suggests that they’re not relying on passive diffusion through membranes for transport. Maybe that’s not important for drugs acting on the outside of bacteria (beta-lactams, vancomycin…) but for agents that have to get to the bacterial cytosol, that probably means they’re relying on fooling bacterial transporters. Small molecule libraries that have been carefully designed for passive transport aren’t going to succeed at that (as repeated HTS efforts have shown).
    Hence the continued dominance of natural products and semi-synthetics.

    1. lynn says:

      Actually, most clinically used antibiotics [even the natural products] are either passively diffused or get in with an energy boost from the proton-motive-force [that allows weakly charged molecules to cross the cytoplasmic membrane) – but are not carrier dependent. Many natural products do, indeed, use transporters – however, only a couple of these (that use transporters) are clinically used because most of the transporters are not essential and so can be lost and lead to rapid resistance. This has not stopped people over the years from trying this route, but, so far, it hasn’t been productive. One exception is fosfomycin which is actively transported – but mutants losing the transport systems do not seem to thrive in vivo.

      1. Barry says:

        thanks, Lyn!
        At least albomycin, and rifamycin are reported to enjoy active transport by FhuA. I don’t know what else might exploit that. They certainly don’t look like the sort of structures that populate most HTS small-molecule libraries.

        1. lynn says:

          Yes for albomycin. Rifamycin itself is not transported, but a semisynthetic derivative, CGP4832, is indeed transported by Fhu/TonB. It has very good activity against some Gram-negatives, but resistance occurs at 5x10e-7 through mutations in fhuA or tonB (Pugsley A, Zimmerman W, Wehrli W. 1987. Microbiology 133:3505-3511).

  23. bacillus says:

    Regarding Phage, they are, like vaccines, very specific so you really need to know what the infecting strain is in order to administer the correct phage. Of course, mixtures of phage could be used, but that would have to be a big mix to guarantee hitting the target. Finally, phages have been responsible for transmitting undesirable properties to normal flora (e.g. E. coli O157:H7 toxin). As for vaccines against MTB, malaria etc, the major problems are that no-one knows whether you need an immune response to just one or to many antigens to elicit protection. Also, you need to generate a cell-mediated response for effective control of the aforementioned, and no vaccine to date has ever been licensed based on their ability to do so. Although some vaccines probably work at least in part by generating cellular immunity the correlates of protection used to license them have been passive or functional antibody assays.

    1. steve says:

      Just a few corrections. First, you don’t need cell-mediated immunity for a vaccine; almost all vaccines work through antibodies and antibodies are perfectly capable of killing bacteria. The issue with vaccines is that you don’t know ahead of time what the cross-reactivity with normal tissue will be so you need to do huge trials. It’s not clear that you’d want to vaccinate against multidrug-resistant E. coli, for example, because you’d wreck havoc on the normal microbiome. Bottom line, you’d never use a vaccine for the normal population as the incidence is so low that you wouldn’t want to vaccinate every normal person and the lag time it takes for immunity is too long to use in an epidemic. You want some sort of orally-active medicine that you can store in case of an epidemic, which leaves you mostly with small molecule drugs.

      1. loupgarous says:

        One example would be umifenovir, an indole-based antiviral which is said to (a) block viral entry into cells and (b) “stimulate the immune response.”

        On point (a) there’s published research (in an Elsevier journal, but what the hell) documenting a possible way umifenovir blocks viral entry into cells.

        Point (b) seems to be because, Russian drug regulations are eerily like what some folks have always wanted FDA approvals to be – “what the hell, sell it over the counter and see who it knocks over… “.

  24. Futurechemistmachinist says:

    History will show that the solution to the problem came from nanorobotics, the same will be true for cancer.

  25. Ed says:

    Following on from DCRogers, I think antibiotic usage in livestock is often overlooked whereas its volume probably exceeds its equivalent in humans, and the concentrations at which the antibiotic is used are much lower in livestock, thus favoring development of resistance. What about reducing usage in livestock to try to save antibiotic classes, in parallel with the initiatives described in the report? Obviously this would be a large economical and political challenge. Should part of the profits made in the agricultural sector be earmarked for development of new antibiotics, aimed exclusively for human use?

    I also think we should consider how this initiative would be managed. What constitutes progress and how much progress per unit of time should be made for such group to continue its pursuit? The multitude of past HTS campaigns was not necessarily driven by our love for HTS, but by the decelerating advances in ensuing lead optimization programs combined with an understandably limited patience on the part of management. I feel many scientists refuse to think about this aspect, but imagine (crowd?) funding this from your own 401k instead of seemingly monetary stream from taxpayers or stockholders.

    1. Ed says:

      ….a seemingly unlimited monetary stream…..

  26. Me says:

    Usage in animals is a different issue, since the antibiotics are put in the feed and hence time above MIC (compliance!) is better, so resistance does not develop nearly as quickly as with humans, who forget doses, and then take 2 to make up etc, so plasma levels fall below MIC and allow resistance to develop.

    there you go – a non-cynical reply.

    1. Ed says:

      To me: My comments were not intended to be cynical.
      My understanding is that antibiotic use in animals is mainly aimed at altering the gut flora to maximize conversion of feed into growth of the animal, and by definition therefore the desired flora will be exposed to sub MIC levels. I don’t recall ever seeing a paper on optimizing exposure of antibiotics in this setting, but admittedly I have not really looked either.

      1. Me says:

        re: cynicism, I’m referring to myself as a one-time frustrated antibac med chemist.

  27. steve says:

    Ed is correct, antibiotics in animal feed is a much greater driver of bacterial resistance than use in humans. About 80 percent of all antibiotics sold in the U.S. are given to poultry and livestock. As a result, multidrug resistant bacteria are now found in meat and produce. The MCR-1 gene is particularly scary, being the first drug resistance gene to polymyxins that’s easily transmittable to common bacteria like E. coli and K. pneumonia.. It arose because colistin (a type of polymyxin) was fed to pigs in China; a recent study showed that a majority of those pigs now have MCR-1. So please don’t take antibiotics given to livestock as a minor problem; it’s actually the major source of antibiotic-resistant bacteria.

  28. I believe the human body makes a non-mutagentic bacterial protein synthesis inhibitor. The presence of 7 billion people argues for the existence of such a compound. I have synthesized what I believe to be such a compound. It shows effectiveness against a wide range of bacteria in non-sterile culture. As an independent researcher, I have carried this project as far as I can by myself. Where do I need to go, or to whom do I need to talk. to make progress on my discovery?

    1. ch says:

      Just post the structure here and one of us will be happy to help you.

  29. mikeb says:

    The other thing that MUST BE ADDRESSED is abuse of pharmaceuticals in non-1st world countries. Have any of you ever been to Asia, Latin/South America, or Africa? Antibiotics are handed out like candy to anyone who walks in. You can get ‘moxy (amoxicillin) in almost any pharmacy in Asia if you simply pretend to have a cough and some sniffles. 1st world nations can develop as many new antibiotics as we want, but the problem of developing resistance is significantly worsened when antibiotics get into the hands of 2nd and 3rd world countries with virtually 0 oversight over pharmaceutical use.

    (and FWIW, 1st world countires are also to blame by doing stupid things like putting huge amounts of antibiotics into live stock.)

  30. matt says:

    So if the barriers to using macrophages means we doubt they can do the job alone, what about using their evolutionarily successful tricks?

    Something like (waving hands furiously), pick a Gram-neg bacteria of interest, search its genome (and colonies) for evidence of viral DNA/RNA, identify the phage, then study how it gains entry to the bacteria of interest. Surely that has been done a lot?

    Is it just that the virus depends on mutating faster and having enough envelope-picking diversity to overcome the bacteria’s defenses (i.e., a trick that non-phage medicine could not hope to do), or are there other tricks? Is there a database of human-pathogenic bacteria, and which viruses are known to target them?

    If it depends on virus-like brute force techniques, could you generate a genetically engineered version of the “natural predator” for your pathogen of interest, which you optimized by selection for the particular environment you want (bacteria specifically in the human bloodstream and its conditions, for example)?

    If you ended up with a solution that was fiendishly difficult and expensive to mass-produce, all the better, right? That way, it would only be used as a last resort. Not gonna work for cows, farmers. Not going to be synthesized cheaply for your sniffles, everybody. Not used prophylactically, but only as a treatment for individuals proven to be infected by that particular organism, and perhaps only for individuals whose infection sequencing indicated that ENV strategy would work. Personalized medicine. Not sure how effective it would be in immuno-compromised individuals.

    I suppose the trick might be in getting your phage to escape the immune system just slightly longer than the bacteria…(left as an exercise for the reader, and all that). And of course that means the next bacterial target will need a different phage substrate, to avoid the likely innate immunity to the first.

    Sorry to go amateur hour like that, but it seems like virology might produce some insights.

  31. Wallace Grommet says:

    Intralytics has been working away at phage therapeutics since founded by former Georgian scientists after funding dried up with the collapse of the Soviet Union.

  32. bacillus says:

    @Steve. You are plain wrong about the need for CMI for certain vaccines to work. BCG and yellow fever vaccine are the classics. Indeed, all live vaccines will almost inevitably produce both antibodies and CMI and that sometimes antibodies alone are sufficient for protection (e.g. anti-toxin vaccines, anti-capsular vaccines). Also, there are many other vaccines in use in which correlates of protection have been developed based on antibody assays (e.g. anti-HA titres in flu), but there is a growing body of evidence that anti-HA titres do not correspond with protection, but having had the assay approved by FDA half a century ago, manufacturers are under no pressure to change it. They still use SRID to measure HA content of flu vaccines, even though much better methods are now available, but again would need to go before the FDA for approval. In the case of BCG, Sanofi Pasteur is now facing a shortage of glass roller bottles for growing it and this is the sole method approved for its production by FDA. Antibody-based correlates of protection are every vaccinologist’s dream since they are relatively straightforward to develop, validate and automate. Remember, that as far as FDA is concerned a correlate of protection does not have to have anything at all to do with the mechanism of protection. All intracellular bacterial pathogens require CMI for protection which is the main reason we don’t have many vaccines against them, and those that we do have are far from perfect.

    1. steve says:

      Not sure where you get that idea. Antibodies are the major immune response to bacteria, not CMI which targets HLA-bound internal peptides on sides.

  33. Mike Robe says:

    Stop with the “Antibiotics in my meat” scare tactics. Consumer goods and silent chemicals trigger resistance too, as does tooth whitener, oral mouthwashes and even fabric softener. So until you know the facts, including that resistance plasmid reservoirs have been around before man and exist EVERYWHERE, start funneling the money wasted on resistance biology into chemical approaches to creating new antibiotics. But leave it to Pew to lament and bombast on the subject, as they do nothing to suggest or support new chemistry.

    And Pharma has no patience for antibiotic chemistry research either, and the few that do are the walking vulnerable in the extinct sense of the word. The experts in chemistry have scattered, leaving the unexperienced to synthesize the next antibiotics- and it’ll be a while before they are up to speed, if their leaders let them.

    1. lynn says:

      Please read the Pew proposal, since it does indeed speak of supporting new chemistry – that is one of its main points. It recognizes that the experts are dispersing and their expertise needs to be captured. Pew does not speak [in this proposal] of antibiotic abuse or agricultural use.

      1. steve says:

        Pew doesn’t but the facts are pretty clear. Sorry, but antibiotic use in agriculture is a much bigger driver of antibiotic resistance in bacteria than fabric softener, as in the example I gave.

  34. Mike Robe says:


    I did read it and the word chemistry was mentioned 20 times in a 47 page document. Not enough. This report relies on the old paradigms, screen this-repository that, -hopefully companies will share data.
    Still wishy-washy. I want to see “We will fund an institute or company that employs chemists to make antibiotics-period!”

    But the problem lies in that antibiotic research is still led by biologists.

    And the problem is to be answered by chemists.

  35. steve says:

    So form a startup company and get to work.

  36. Mike Robe says:

    That doesn’t make any difference, startups and even delivering actives to corporate alliances. Unless you have been through the process-discovering an antibiotic through to the clinic- you will never see through the facade/snake pit that is antibiotic drug discovery.

    1. Lyle Langley says:

      Mike Robe: I want to see “We will fund an institute or company that employs chemists to make antibiotics-period!”
      Steve: So form a startup company and get to work.
      Mike Robe: That doesn’t make any difference, startups and even delivering actives to corporate alliances.

      Hmmm…. A little bit of everything Mike Robe. As Steve suggested, if you want someone to fund a company employing chemists then go on with your bad self and start one, rather than complaining about what the Pew isn’t doing.

      1. Derek Lowe the pitcher says:

        Could I buy a monorail?

      2. Derek Lowe the pitcher says:

        Lol, that’s probably your real name. Not Lyle Lanley.

        1. Lyle Langley says:

          I wish I were Lyle Lanley. Working over those rubes from Springfield.

  37. Bagnar says:

    Thank you to all and especially you Derek, from bringing this topic here.
    I think antibiotics are more than ever needed and this area need more than steriles discussions.

    I found some really interesting papers in these comments, thank you all 😉

  38. bacillus says:

    @Steve. I’ve been in the vaccine game for 34 years. Your insistance that antibodies are all that are needed to kill bacteria suggests you haven’t been in the game at all. As I said, sub-unit vaccines that elicit antibodies that neutralize toxins or bind to polysaccharide capsules are all that are needed to combat some pathogens, but not all. How the hell do antibodies get to intracellular bacteria? Thirty seconds on Google Scholar would disabuse you of your simplistic grasp of how vaccines work. Just because vaccines produce antibodies doesn’t mean that antibodies are responsible for the observed protection. BCG (over 1 billion doses administered and counting) produces antibodies, but they do sod all to control the growth of MTB. Vaccinia induces antibodies that kill Variola in vitro, but they also produce T cells that can kill the latter virus in vitro too. The relative contributions of each to eliminating smallpopx from the world remains contentious to this day.

    1. steve says:

      The major mdr bacteria causing the problems today are not intracellular bacteria. Sure you need cellular immunity if you’re going to kill chlamydia. But Vancomycin-Resistant Enterococci (VRE), Methicillin-Resistant Staphylococcus aureus (MRSA), Extended-spectrum β-lactamase (ESBLs) producing Gram-negative bacteria. Klebsiella pneumoniae carbapenemase (KPC) producing Gram-negatives are all extracellular and fully exposed to antibody responses.

  39. steve says:

    BTW, you’ll have to explain to me how cellular immune responses (by which I assume you mean T cell) kill extracelluar bacteria when there is no HLA-bound peptide for the T cell receptor to recognize. There must be some new biology there which you should publish. I think you’re conflating results obtained with viruses (which are indeed intracellular) with extracellular bacteria.

  40. Demi says:

    One approach would be to find something so essential to bacterial survival that resistance is not possible — every mutant (within a reasonable genetic distance from what is already present) that would be resistant is non-viable. Maybe something based on lysozyme (for Gram-positives)?

    Or maybe the real solution is to find something that is effective against something found in all bacteria. All bacteria are prokaryotes. Humans are eukaryotes. Perhaps one could exploit this.

  41. loupgarous says: shows that FDA (at long last) is addressing the activity of one very active constituency for MDR bacteria – agribusiness.

    FDA is commendably removing “promoting growth” from the labels of antibiotics “medically important to humans”, and requiring veterinarians to supervise the addition of these antibiotics to humans. According to the author of this Pew trusts report, the drug manufacturers are voluntarily accepting this guidance.

    However, the new FDA policy creates a loophole as big as one of Frank Perdue’s chicken farms – it only applies to antibiotics “medically important to humans”. Anyone familiar with how agribusiness works knows that this policy can be circumvented by taking an antibiotic that isn’t used in humans, but is chemically similar to ones that are – the tetracyclines in particular are used without any limit to duration of use in food animals.

    In too many cases, a plasmid or other resistance mechanism which confers resistance to a single antibiotic will confer resistance to many chemically-related antibiotics. The tetracyclines, for example. So FDA needs to consider tackling the MDR bacteria lobby in Congress, agribusiness, and ending the use of any antibiotic which can cause cross-resistance in bacterial strains to medically-important antibiotics.

Comments are closed.