Intelligent Design

Rubisco is not an example of unintelligent design

Spread the love

The claim that Rubisco is poorly designed or unintelligently designed was appearing in textbooks in the 1990s. The idea has been picked up recently in a News & Views piece by John Ellis. He writes that Rubisco “is a relic of a bygone age” and his essay has the title: “Tackling unintelligent design”.

“Rubisco is the most important enyzme on the planet – virtually all the organic carbon in the biosphere derives ultimately from the carbon dioxide that this enzyme fixes from the atmosphere. But Rubisco is also one of the most inefficient enzymes on the planet. It evolved when the atmospheric composition was different from that of today, and its failure to adapt significantly to the modern atmosphere limits agricultural productivity.”

Over the past decade, the pendulum has swung away from the idea that Rubisco is unintelligent design. [snip]

[After drawing attention to some significant papers, the blog continues: ]

Design theorists have drawn attention to three additional considerations:
1. A single-factor analysis of Rubisco is inadequate. The parameters considered to conclude the enzyme is poorly designed and inefficient are very limited. We should note that our perceptions of intelligent design are typically subjective, and most claims for poor design do not stand up to the test of time – further research leads to a greater appreciation of design (a good example being mammalian eye design). Furthermore, unintelligent design of architectures we deem sub-optimal should not be regarded as the only possible hypothesis. Multiple factors are likely to be relevant as chemosynthetic carbon fixation also makes use of Rubisco. It is employed by organisms living at hydrothermal vents and cold hydrocarbon seeps.
2. Photorespiration, the consumption of oxygen to produce a sugar that ultimately forms carbon dioxide during a series of reactions, may not be a mark of inefficiency, but the process may be useful to the plant. The null hypothesis for Design theorists is that processes have functionality. This hypothesis is not without some support: the process of photosynthesis is not just to capture CO2 and release oxygen because nitrate assimilation in plant shoots depends on photorespiration, as Rachmilevitch et al (2004) have shown.
3. Ecological considerations should be included in the analysis. If design is relevant to understanding the way plants work, we should consider not only the benefits to the organism (which limits the horizon for those with a Darwinian perspective) but also the biosphere as a whole. Rubisco’s ability to capture CO2 increases with increasing CO2 content in the atmosphere, so its efficiency rises in a CO2-rich atmosphere. However, increasing oxygen levels in the atmosphere will reduce Rubisco’s ability to capture carbon. So a negative feedback mechanism exists to regulate the relative concentrations of oxygen and carbon dioxide in the atmosphere. This is another example of design affecting the Earth’s ecology – for more on this, go here.

Ellis was commenting on a paper by Liu et al. that reports on work to produce a Rubisco in vitro. In order to do this, the authors required two chaperone proteins, ATP and addition of 18 protein subunits (taken from a cyanobacterial Rubisco) to be introduced in the correct sequence to get yields of the enzyme. It is hoped that this procedure can be used to produce mutated versions that can be screened for improved effectiveness. It’s all very interesting, but the biggest mystery is why people who expend so much intellectual energy on improving this remarkable molecule can live with the thought that “Rubisco is a superb example of unintelligent design for the modern world”. Maybe research funds would be better spent exploring avenues identified using the presumption that this enzyme is optimally designed.

This issue has been noted in an earlier blog on UC, but additional perspectives are worth bringing to the table. For more, go here.

8 Replies to “Rubisco is not an example of unintelligent design

  1. 1
    David Tyler says:

    This quotation was sent to me by a UD commentator. I reproduce it here because it is directly relevant to the point (above) about judgments based on single factor analyses or other limited perspectives about Rubisco functionality:

    “If the arrangement of an organism seems functionally inappropriate, the most likely explanation (by the test of experience) is a faulty view of its functioning.”
    Steven Vogel (2003): Comparative Biomechanics . Princeton University Press, p15.

  2. 2
    Sooner Emeritus says:

    All biological evolution is coevolution.

    If there is any sense in regarding biological evolution as optimization (that’s a big if), then it is as multi-criterion optimization. There is no practical way for us to identify all of the criteria in study of past, let alone future, evolution. Thus the only way to argue for the optimality of a biological system is to impose a measure of performance. It is important to remember that WE define the measure, and that nature does not offer it up to the observer.

    A huge problem in engineering is that the performance of deployed systems degrades as operating environments change. Thus engineers have mimicked biological evolution in design of adaptive systems. These evolving systems sometimes achieve high levels of engineer-defined fitness with behavior the engineer neither intended nor desired. For instance, a genetic algorithm working in combination with given software for financial modeling has yielded individuals of high fitness that exploit bugs in the given software. I believe that biological evolution similarly exploits opportunities in the environment that seem “weird” to us.

    My point is that there is good reason to design a system to adapt to changes in its environment, and that the subsequent trajectory of the INTERACTING system and environment is generally unpredictable over the long term. That one can define a performance criterion in the present and say that a system is suboptimal in performance is utterly irrelevant to the question of whether the system was designed.

    There seems to be little appreciation here of the difference between design and control. There is no reason to believe that a system designed for multi-criterion optimality long ago should be optimal in the present, unless you believe that the designer knew the future, and was capable of designing for optimality at all points in time. Engineers certainly do not have this capability. If a system is not adapting to a changing environment, then its enduring optimality in performance is due to control of the system, not design.

  3. 3
    Sooner Emeritus says:

    If a system is not adapting to a changing environment, then its enduring optimality in performance is due to control of the system, not design.

    Clarification: There are two ways (not mutually exclusive, actually) for a system to remain approximately optimal in a dynamic environment. The system uses some sort of feedback to adapt to the environment or the an external entity applies control inputs to the system.

    Intelligent design fails as a theory if it does not bound the knowledge and capabilities of designers. A designer cannot know everything about the future, and cannot know how to design a system that is optimal in an absolute sense now and forever more.

    In other words, omniscient and omnipotent agents certainly have no place in scientific explanations.

  4. 4
    Oramus says:

    David,

    FYI, you’re missing a link at the end of no. 3. Could you add it, please?

    Thx.

  5. 5
    David Tyler says:

    Oramus @ 4
    Done.
    Thanks for spotting the omission.

  6. 6
    Joseph says:

    Is it just me or does “Rubisco” sound like an entree from high-end restaurant?

    Come to think about it it could also be some thrown together grub from some local eatery.

    Or perhaps a sauce- would you like some rubisco on that?

    Or an ointment- rub some rubisco on that and it will clear up.

    Fun with biology!!!

  7. 7
    Sooner Emeritus says:

    Yes, Rubisco rhymes with Nabisco. Fun.

    1. What does it mean to say that a biological system is optimal in design, when we cannot know the objectives of the designer?

    2. How can a designer with bounded knowledge and capability design a biological system that is sure to be optimal in an absolute and nontrivial sense over the long term?

    Suppose that Rubisco was optimal in design long ago, according to someone’s criterion of optimality. This does not imply that Rubisco should be optimal today by the same criterion. And why would we talk about this anyway, considering that we cannot know the criterion?

    Don’t make me invoke Wittgenstein! 🙂

  8. 8
    Collin says:

    Sooner Emeritus,

    I think those are good questions. Maybe it is not appropriate to say optimal design.

    Perhaps a better endeavor would be like this: we first assume that the function that Rubisco performs is the one that it was meant to perform. This might not be true, but is a good starting point. From there see if there are better or worse ways of doing it. We won’t find what is optimal per se, but we will be able to rate the relative “optimaliability” of it. And then we could look for systems that are more likely in a design scenario and ways that are more likely in an undirected scenario.

Leave a Reply