Flagellar Variations, Minimal Complexity, and Evolutionary Noise.


This article will seek to accomplish two things. First, it will re-emphasize the IC state of eubacterial flagella despite the variation that is seen in such structures. Secondly, it will outline ways in which a design perspective can guide research and even make predictions *about* evolution. This article does not intend to convince anyone that design has occurred or that the flagellum is the product of design. This article is written for those who are open to both design and evolution, who suspect there is some kind of truth to the design inference, and who are thus curious as how design concepts can be used to guide research.

The Varying Flagella and Sloppy Simplicity

The bacterial flagellum remains a strong candidate for a design event. Not only does a comparative analysis demonstrate Ur-IC without any evidence of an evolutionary history (a pattern that is predicted from a de novo design event), but the structure so strongly resembles a designed artifact (a propellar driven by a rotary motor) that several biologists have commented on this. In my previous article on the design of the flagellum, I outlined the specific components that are part of the E.coli flagella and demonstrated their irreducible complex nature not only in terms of structure/function, but also in terms of synthesis.

Although the E.coli flagellum, with about 40 flagella proteins, shows irreducible complexity and the basic architecture of the E. coli flagellum is widely distributed among eubacteria, the sequencing of recent genomes have shown that flagellar function can occur with a smaller number of components than seen in E. coli. For example, biochemist David Ussery criticizes Behe’s example of the flagellum by pointing out the following on his web site:

“However, in the bacterium that causes syphilis (Treponema pallidum), there are a total of 38 flagellar proteins; in the bacterium that causes lyme disease (Borrelia burgdorferi), there are only 35 flagellar proteins; finally, in a bacteria associated with ulcers (Helicobacter pylori) there are only 33 proteins necessary to form complete, fully functional flagella. It is likely that as new bacterial genomes continue to be sequenced (at the rate of about one a month!), organisms will be found which require even fewer genes to make a completely functional flagella.”
I only counted 36 flagellar genes for T. pallidum. And Aquifex aeolicus (a bacterium not mentioned by Ussery) only has 27. This would seem to supports Ussery’s contention. But does it?

Ussery’s point is interesting not as any type of refutation of IC, but as a launching point for further inquiry employing the dual concepts of systematic and thematic IC (see my earlier postings “DNA replication and IC” and “Design and the F-ATPase” for explanations of these concepts). In the examples cited by Ussery the number of systematically IC components ranges from 33-40 players. This, of course, is not a terribly impressive variation when we consider that spirocheates and proteobacteria are distantly related. In fact, one might have expected a larger variation with reference to spirocheates given that they employ their flagella in a novel means of locomotion.

Nevertheless, it is apparent how the notion of sloppy simplicity is guiding Ussery. The response to IC is to identify simpler and simpler systems with the hope that the IC essence will disappear. This same logic is used elsewhere in the same article by Ussery:

“If you look at bacterial flagella, you find that some are indeed quite complicated, but others are more simple. For example, the basal body can vary with species – in E. coli there are four rings, in Bacillus subtilis two rings, and in Caulobacter crescentus five rings. I can easily imagine a scenario where a “primitive bacterium” might have one ring, and then you have a flagellum with two rings, then three, and so on. This is a “gradual, step-by-step” evolution, which is the antithesis of Behe’s argument.”

While there is variation among flagella we need to inquire as to the nature of this variation (Ussery’s imagination is irrelevant). This variation could be “large-scale” where significant thematic components are lost and replaced by novel components. Or as Ussery implies, does this variation reflect simpler and simpler states? Such large-scale change would indeed pose a challenge to the IC essence of the flagellum. On the other hand, the variation may be “small-scale” where peripheral parts have been modified, lost, and added by evolution. This would constitute “evolutionary noise” layered on top of a design event as evidenced by an IC core. This change is consistent with my hypothesis of design followed by a history of evolution (my critics continually err in thinking that evolution has no place in my hypotheses). To determine what type of change is evidenced by this variation, we need to determine first if the eubacterial flagellum possesses an IC core, and if so, what it is.

The Flagellar IC Core

To determine if the eubacterial flagellum has an IC core, the flagellar genes from four species of bacteria were compared. These included the gm+ soil bacterium, B. subtilis; the spirochete, T. pallidum; a gm- proteobacterium, H. pylori; and the thermophilic, gm- bacterium, A. aeolicus (this genus represents the deepest branching lineage in the eubacterial tree). These four were chosen because their genomes have been sequenced and because they are all distantly related from each other, so much so that it is possible to infer the flagellar state in the last common ancestor of all eubacteria.

All structural flagellar genes were tabulated for each species (B. subtilis has recently duplicated many of its flagellar genes and these paralogs were not included in this analysis). From my scoring, there are 18 genes shared by all four species. They are listed below:

fliS, flgE, flhA, flhB, fliR, fliQ, fliP, fliI, motB, motA, fliD, fliG, flgB, flgC, fliF, flgK, fliC, and flhF.

These genes apparently represent an IC core. To determine how
well this IC core maps to a designed artifact, the functions of each gene were determined to see if they would fall nicely into thematic IC (where components and sets of components would fall into several functional themes). The demonstration of thematic IC can be viewed as the “design blueprint” for making the bacterial flagellum.

As it turns out, most of these core IC genes fit into several thematic themes as shown in Table I.

Table I. Flagellar Components Appear to Demonstrate Thematic IC.

Theme Genes Components
Capped filament fliD, fliC/flaA/flaB*
Hook and adaptors for attachment to filament flgE, flgK
Drive shaft flgB, flgC
Motor complex (MS ring complex) motA,motB, fliF
Switch fliG
Export machinery flhA,flhB, fliR, fliQ, fliP,, fliI
Unknown fliS – essential, but function unknown
flhF – GTP-binding protein

*C,A,and B are different variants of flagellin filament proteins and
are thus grouped.

Many interesting insights emerge from Table I. First, flagellar stucture and function in eubacteria appears to depend on 6 different functional themes. However, much more research is needed. On one hand, the switch and motor complex may be viewed as one functional theme. Yet on the other hand, the motor complex may be viewed as involving a stator and rotary motor theme. Furthermore, do fliS and flhF constitute new functional themes or do they fit into pre-existing themes? A design theorist would be interested in a better understanding of these functional components to more clearly outline the individual functional themes that are part of thematic IC. Clearly this is another area where design could guide scientific research.

Secondly, as with the F-ATPase, the functional themes appear to be comprised of more than one systematic component. Thus, once again, we see how systematic IC can be used to infer thematic IC which can, in turn, be used to explore systematic IC in a more focused manner. The exception appears to involve the switching function (components employed to switch flagellar rotation from clockwise to counterclockwise and back) which involves in one gene product. This will be discussed below.

Finally, when we consider the manner in which the IC core reflects thematic IC, the variation noted by Ussery is not that impressive. That is, each theme appears to be essential for flagellar function and in no case do we find the variation any of them to be omitted. Thus, the variation we see does not demonstrate variation in the actual components that represent the themes (in other words, there is no other motor complex other than the MS ring complex), but instead represents tinkering around the functional themes.

Minimal Complexity of the Bacterial Flagellum

Thematic IC suggests that six independent functional themes are needed for flagellar function in eubacteria. These six themes are represented by 18 indvidual components. Thus, these 18 components reflect thematic-systematic IC(t-s-IC). But does t-s-IC represent the minimal complexity needed for flagellar function? This seems unlikely as no flagella have been found with these few components. H. pylori has 33 components, T. pallidum has 36, B. subtlis has 31, and A. aoelicus has 27. That the simplest flagellum (A.aeolicus) contains nine more components than s-t-IC suggests that s-t-IC represents only the minimal shared complexity that arose from the common ancestral state (ur-IC). Thus, the Ur-IC state would include these 18 gene products, however, this Ur-IC state may not reflect the minimal complexity needed for flagellar function. Put simply, to make a functioning flagellum, we need not only these 18 gene products, but some other undetermined products whose identification has been blurred by evolution.

To estimate the minimal complexity of the Ur-flagellum, further scoring was done with the four model species (redundant gene duplications were not common and also not counted).

Table II. Genes found in 3/4 lineages.

Gene Species missing from Function
fliL H. pylori flagella assembly
fliH A. aeolicus flagella assembly
fliY A. aeolicus switch
fliM A. aeolicus switch
flgL H. pylori hook assembly
fliE A. aeolicus hook adaptor prot
flgG B. subtilis rod protein

Table III. Flagellar genes found in 2/4 species

Gene Species with Gene
flgD T. pallidum/H. pylori
flgH A. aeolicus/H. pylori
flgI A. aeolicus/H. pylori
fliN A. aeolicus/H. pylori

Table IV. Species-specific flagellar genes

Species Genes
T. pallidum flbD, cpfA
A. aeolicus flgA
B. subtilis fliZ, flhD, flhP, fliJ, fliK, fliT
H. pylori flaG

To estimate the minimal complexity, individual flagellar components were weighted by their relative incidence and summed (species-specific genes were averaged (2.5 species-specific genes per lineage)).

18 (1.0) + 7 (0.75) + 4 (0.5) + 2.5 (0.25) = 25.87

This figure suggests that a minimum of 26 genes products are needed to form a functional eubacterial flagellum. That this number maps closely to the number of components to A. aeolicus suggests this species cannot simplify its flagellum much further (furthermore, the 27 genes of A. aeolicus should not be thought to represent all the 26 gene products that would be part of the Ur-IC state).

But what if we made the same calculations with only three of the genomes? If we remove A. aeolicus from the comparison, I get a minimal complexity value of 25.97. If we remove B. subtilis, the minimal complexity becomes 26.10. If H. pylori is removed, the minimal complexity is 24.77. If T. palladium is removed, the minimal complexity is 25.66. Thus, by looking at all four permutations of three of the four genomes, the minimal complexity ranges from 24.77-26.10 with an average value of 25.62.

Thus, it appears that a minimum of 26 protein products are needed to construct a functioning bacterial flagellum. We can propose therefore that the designed flagellum was constructed with at least 26 proteins, although it may have had more, as auxiliary proteins may have been part of the original design. If these 26 proteins constituted an IC system, then to evolve this system, 26 proteins with similar activities, yet part of an undefined number of different systems, had to exist and then all disappear without a trace, leaving behind only the bacterial flagellum.

Thus far, we can predict that the originally designed flagellum was composed of approximately 26 gene products and that this minimal complexity is needed to form a functional flagellum. Of those 26, 18 are still identifiable as part of the Ur-IC state. A simple prediction thus follows that is easily testable. While Ussery predicted more genomes will contain fewer and fewer flagellar genes, I predict that one will not find a functioning flagellum with much less than 26 genes. I further predict that that the Ur-IC state (or most of it) will be seen in future genomes of flagellated bacteria.

Evolutionary Noise

Both Ur-IC and the minimal complexity also allows the design theorist to identify the degree of evolutionary noise (EN) that has occurred since the design event. To determine EN, one first determines the number of proteins that have been lost, replaced, or added since the design event by subtracting the identifiable Ur-IC state from the min-complexity, 26 – 18 = 8.
The evolutionary noise coefficient (ENC) is thus determined
as follows:

ENC = # genes lost, replaced, added by evolution
Ur-IC + # genes lost, replaced, added by evolution

ENC = 8/ 18+8 = 0.308

The ENC allows the design theorist to estimate the degree of plasticity inherent in the designed system. For example, a novel system that evolved in a specific species would not have any shared components in other species. This system would not demonstrate systematic-thematic IC and thus an Ur-IC state would not be identified (Ur-IC = 0). In this case:

ENC = # genes lost, replaced, added by evolution
# genes lost, replaced, added by evolution

ENC = 1

If, however, no evolutionary change has occurred in the system, the # genes lost, replaced, and added by evolution would equal 0. The ENC = 0. Thus, the closer the ENC is to 0 the more the system has resisted change and the closer we are to the original designed state. Further analyses with other IC systems are needed to determine how significant the change is that is represented by an ENC of 0.308.

Reassessing the Scoring in Light of Thematic IC.

That six flagellar genes are found in 3/4 distantly related bacteria raises the possibility that thematic IC might help us identify further specific components of the Ur-IC state. There are two noteworthy examples.

The Bushings

The first thing to note as the something obvious is missing from the thematic list, namely, the L and P rings that form bushings in E. coli. This lack is due to the inclusion of B. subtilis in the comparative analysis. B. subtilis is a gm+ bacterium and thus has no need for P and L rings. Should the P and L rings be included as part of the Ur-IC state? Peter Nykios has proposed on talk.origins that the gm- state is more ancient than the gm+ state. If this is the case, the original designed bacterium would be gm- and require L and P rings. As support of for this hypothesis, A. aeolicus, which not only has the simplest flagellum but also branches deepest in the eubacterial phylogenetic tree, has both L and P rings (from the genes flgH and flgI, respectively) as it is a true gm- bacterium. In addition, A. aeolicus also expresses flgG (as do T.pallidum and H. pylori, but not B. subtilis). One could thus propose, as Peter did, that the gm+ state evolved from the gm- state early in evolution.

It is important to understand that a design perspective is not an anti-evolutionary perspective. That is, both Peter and I propose design events that are followed by evolution. What this means is that the design perspective Peter and I share allows us to view *evolution* in a manner that can be quite different from the manner it is commonly viewed. In this case, a design perspective allowed Peter to predict that the gm+ state evolved from the gm- state. His critics, lacking his design perspective, were unable to make any type of prediction about this. Thus, design can lead to predictions about evolution (and in some cases, these predictions depend on the design perspective).

To support his prediction, Peter proposes that is far easier for evolution to lose flagella than acquire flagella. There is good support for this view, namely, we can track the loss of flagella. For example, Shigella boydii and Shigella sonnei are two species of bacteria without flagella. However, flagellar genes were found in both strains containing several deletions and disruptions (Al Mamun, et al., 1997, Cloning and characterization of the region III flagellar operons of the four Shigella subgroups: genetic defects that cause loss of flagella of Shigella boydii and Shigella sonnei. J. Bacteriol. 179, 4493-500). It is thus likely that flagella have been lost independently on multiple occasions when bacteria occupy a niche where flagella are not longer needed.

But what we don’t see are any bacteria in the process of acquiring flagella. This is odd as species which have lost flagella long ago might find themselves in environments where flagella would again be favored. But this is not surprising given the IC nature of flagella. Unless all genes are present to form a flagellum, they would have no function and decay like pseudogenes (as in Shigella).

In fact, once we realize that flagella have been lost in many lineages and that some of these non-flagellated bacteria may have later found themselves in environments where a flagellum would be advantageous, we can see more clearly why the non-design explanations are weak. Put simply, where are the different flagella themes within eubacteria? Consider that Howard Hershey is arguing that the flagella of eubacteria and archeabacteria show the independent evolution of such structures (I look at Howard’s hypothesis in another posting). Well, why don’t we see the independent acquisition of flagella *within* the eubacterial tree? If flagella can be independently acquired by evolution, why have they not been independently re-acquired in eubacteria? Why haven’t any eubacteria that have lost their flagella given rise to a lineage that has re-evolved a *different* type of flagellum?

To summarize, the loss of flagella are easier to evolve than the acquisition of flagella. This view is not only supported by empirical evidence (where we see flagella in the state of being lost but not in the state of being acquired), but also by the multiple loss-events giving rise to not a single re-acquisition of an independently evolved flagellum. With these points in mind, I would have to agree with Peter and tentatively propose the gm- flagellum, with both P and L rings, is closer to the originally designed flagellum in the first bacteria (this proposition assumes, however, that only one form of flagellum was designed in eubacteria).

Support for this hypothesis not only stems from the fact the A. aeolicus has a flagellum with both P and L rings, but it looks as if the rings are in the process of being lost in spriochetes. Sprirochetes are considered to be analogous to gm- bacteria in terms of their cell membrane structure, yet their flagella are uniquely arranged to form axial filaments. The genome of T. pallidum indicates that it possesses neither the P or the L ring (flgI and H, respectively). But the drive shaft still incorporates the flgG protein which is where the L and P ring are attached in gm- bacteria. The presence of flgG may be vestigial in spirochetes reflecting that they once had L and P rings. To test this hypothesis, I surveyed the genome of another closely related spirochete, B. burgdorferi. Like T. pallidum, it still has the flgG protein. However, B. burgdorferi also has the P ring (flgI), but not the L ring. T. pallidum and B. burgdorferi thus look like bacteria in the process of changing a gm- flagellum to something that looks more like a gm+

The Switch

Another interesting aspect of Table I is that the switch is represented by only fliG. This is significant as there is extentive data to indicate the fliG works as part of a complex with fliN and fliM. Comparing the four genomes, it becomes clear that in addition to fliG, *either* fliN *or* fliM is present. For example, while fliM is lacking in A. aeolicus, it possesses fliN. And while fliN is lacking in B. subtlis and T. pallidum, both contain fliM. Thus, a thematic perspective may group fliM and fliN together where at least one of these two components is needed as part of the switch complex with fliG.

What is interesting about this evolutionary loss is that it actually highlights how an IC complex may be composed of IC subsystems. FliM is lacking only in A. aeolicus. Plenty of data exist that suggest FliM’s role in the switch complex is to act as a mediary between the esential FliG component and the chemotaxis protein, CheY (for example, see Toker and Mcnab, 1997. JMB 273; 623-634). If the coupling of chemotaxis to flagellar rotation is through a mini-IC system, one would expect that without FliM, CheY could not function. And indeed, in A. aeolicus, not only is FliM lacking, but so to is the entire chemotaxis system including CheY. All of this suggests that A. aeolicus lost the need for chemotaxis and thus lost all the genes that are part of this potential mini-IC system.

If this analysis is valid, we can modify Table I to include one new functional theme with three components and add an additional component to the switch.

Table V. Modified Thematic IC Scoring

Theme Genes/Components
Capped Filaments fliD, fliC
Hook and adaptors for attachment to filament flgE, flgK
Drive shaft flgB, flgC
Bushings/bearings flgG, flgH, flgI
Motor complex (MS ring complex) motA,motB, fliF
Switch fliG, fliM/N
Export machinery flhA. flhB, fliR, fliQ, fliP,, fliI
Unknown fliS – essential, but function unknown
flhF – GTP-binding protein

Thus, we can raised our Ur-IC scoring from 18 to 22 suggesting that these 22 gene products were present in the LCA of eubacteria. But what does this do to our scoring of minimal complexity?

Min complexity = 22 (1.0) + 5 (0.75) + 1 (0.5) + 2.5 (0.25) = 26.87

By using evolution and thematic IC to shift 5 gene products into the Ur-IC state,the minimal complexity has increased from 26 to 27 genes. This maps perfectly to the number of genes seen in A. aeolicus and again suggests that this ancient lineage of bacteria expresses a flagellar state that is close to the minimal functional state.

This shift also causes the evolutionary noise coefficient to drop:

ENC = 5/22 + 5 = 0.185

But how does the evolutionary noise coefficient drop by incorporating evolutionary events? The key is that the ENC represents unidentified evolutionary noise that could be used to dislodge the design inference. The evolutionary events identified above do not in anyway cause us to reconsider the scoring of the design inference. Less noise means there is less likelihood that future understanding about the evolution in these systems will cause us to reassess the design scoring.

Homologous genes?

As seen in Table V, the flagellar protein export machinery is composed of a sub-IC system with 5 genes, flhA. flhB, fliR, fliQ, fliP,, fliI. Most of these genes have homologs in other systems. FlhB, fliR, Q, and P have homologs in the type III secretory system. FliI has a homolog with the beta-subunit of the F-ATPases. None of these homologs, however, provide good evidence for the step-by-step evolution of the flagellum.

The sequence similarity between FliI and the beta subunit of the F-ATPase is limited and may reflect only the shared ATP-binding domains of both proteins. No other component of the F-ATPase and flagellum share any sequence similarity.

As proposed earlier, the type III secretory homologs probably exist as a consequence of the type III system evolving *from* this export machinery of the flagella after gene duplication (and possibly horizontal transfer). As noted above, design allows one to propose specific hypothesis *about evolution.* In this case, if the flagellum was designed, it is predicted that the type III system evolved *from* flagellar components. This prediction is clearly supported by the fact that not only are these export components part of the Ur-IC state, none of the four species analyzed have the type III secretory pathway.


I have attempted to show how variation within in an IC system does not weaken the design inference that follows from the IC system. On the contrary, it has always been granted that a history of evolution would follow a design event. Thus, the perspective that employs design also brings perspective on how evolution has occurred since the design (showing that design is NOT “anti-evolution”). In this case, an IC core of 18-22 components have been identified that probably reflects a significant fraction of the Ur-state (the designed state in the last common ancestor of eubacteria). The variation that is seen since this design event does not remove any of the six functional themes or significantly change the components involved in any functional theme.. The minimal complexity of a functioning eubacterial flagellum is predicted to be approximately 26-27 proteins, where 18-22 have remained to reflect the designed state. This translates as an evolutionary noise coefficient of 0.185 – 0.308. The evolutionary noise coefficient does not represent any form of argument against evolution or for design, it only serves as a tool for the design theorist to estimate the degree to which the system she studies today reflects the originally designed system. In the end, we are still left with the flagellum as an IC system and a way to explore its design.

Finally, this form of analysis depends on the existence of genomes that have been entirely sequenced and genetic and biochemical data to assign functions to gene products. Thus, a design perspective would clearly guide research in terms of traditional science. A design perspective would encourage that more genomes be sequenced and more genetic and biochemical studies be conducted. That the design perspective may not be needed is an irrelevant point.

Appendix – Identifying Minimal Complexity and ENC in other IC Systems

To further test the notion of minimal complexity and evolutionary noise, the same analyses done with the flagella above were done on all three of the other IC systems I have scored as design products.


The F-ATPase has been previously analyzed in detail to find four functional themes, each with two systematic components.
H. pylori, A. aeolicus, and B. subtilis were compared as spirochetes have replaced their F-ATPases with the V-ATPases (possibly through horizontal transfer).
All 3 species have the alpha, beta, gamma, delta, episilon, a, b, and c

In both H. pylori and A. aeolicus, there is a duplicate for the b subunit, but redundant duplicates are not counted. B. subtilis has a species-specific factor coding for an I subunit as part of the F0 complex.
Thus, minimal complexity is:

8 (1.0) + 1(0.33) = 8.33.

The minimal complexity is slightly higher than the number of components discussed in my article, “Design and the F-ATPases.” It is predicted that further comparative analyses among other eubacterial lineages will decrease this value. The genome of Mycoplasma has been sequenced and represents the smallest cellular genome known. Mycoplasma encodes for the F-ATPase and has 8 subunits.

Evolutionary noise = 8.33 – 8 = 0.33

ENC = 0.33/ 8+0.33 = 0.04.

The ENC among the F-ATPases is close to 0 and thus suggests that this modern complex closely resembles the originally designed state.

DNA Replication

The components of DNA replication were scored in the same four species in which flagellar components were determined. However, at this point, the systematic IC components were not collapsed into thematic IC through a more detailed functional analysis. Also, genes involved in DNA modification and recombination were not included in this analysis.
Four of the species contained 11 of the same gene products, three shared 4 gene products, two shared six gene products, and there was an average of 2 species-specific gene products.
Min comp = 11 (1.0) + 4 (0.75) + 6 (0.5) + 2 (0.25) = 17.5

To test this score, the number of gene products employed in Mycoplasma DNA replication was conducted afterwards. When this value is compared to the number of DNA replication proteins in Mycoplasma (which represents the simplest cellular genome known), the discrepancy is small (min comp = 17.5 and Mycoplasma = 16). Furthermore, Mycoplasma contains all of the 11 gene products that are scored as part of the Ur-IC system. Further analysis of the functional roles of non-universal gene products may serve to lower the min comp score.

The ENC was calculated:

ENC = 6.5/11+6.5 = 0.371

This value suggests that more evolutionary noise has obscured the originally designed state of the DNA replication machinery than either the F-ATPase or flagellum. However, future analyses in light of thematic IC may also reduce this value.
To test this idea, I surveyed the genomes of H. influenzae and the cyanobacterium, Synechocystis. In this case, the minimal complexity was calculated to be:
11 (1.0) + 2 (0.833) + 1 (0.666) + 2 (0.5) + 4 (0.333) + 2.16 (0.133)
= 16.01.

Since the Ur-IC score remained unchanged (11), the ENC
dropped to 0. 313.

Thus, by comparing more genomes, not only does the minimal complexity value map precisely to the number of gene products used in M. genitalium, but the ENC value drops 15%. A major caveat in this analysis is that replication genes are usually grouped with repair, recombination, and modification genes and I excluded these from this analysis. Thus, it is possible that some of these genes play roles in replication and the scoring may be off significantly. In the future, I hope to revisit this whole analysis in a more rigorous fashion.

The Peptidoglycan Cell Wall

An earlier analysis of the bacterial cell wall demonstrated the IC nature of peptidoglycan synthesis. The same four species were again compared, but like DNA replication, no effort was made to collapse the systematic components into thematic IC.

All species shared 12 gene products, three of the four shared one gene product, two of the four shared 4 gene products, and an average of 4.75 gene products were species-specific.
The min complexity involved in peptidoglycan synthesis is
as follows:

Min comp = 12 (1.0) + 1 (0.75) + 4 (0.5) + 4.75 (.25) = 15.94.

Thus, 16 gene products are apparently required to synthesize the cell wall, twelve of which can be assigned to the original state. In this case, it was not possible to check this number against the genome of Mycoplasma as this genus is unique among eubacteria in having no cell wall.
The evolutionary noise coefficient was determined to be 0.24 showing that all four systems I have discussed have similar ENCs (see Table VI).

Table VI. The Evolutionary Noise Coefficients of Four IC Systems in Distantly Related Eubacteria

System ENC ENC after further thematic analysis
Flagellum 0.308 0.185
F-ATPase 0.04 0.04
DNA rep 0.314 ND
Cell wall syn 0.241 ND

These data clearly suggest that the F-ATPases may represent the system that is closest to its originally designed state. This further suggests that this system may be the best system for study by a design theorist and supports my attempt to decipher design motifs (ie, NAIGS) within this particular system (see “Design and the F-ATPases”).

Other suggestions for future research and analyses include:

–Sequencing of more genomes and re-scoring the Ur-IC state, minimal complexity, and ENC in light of this new information.

–Discovery of more functional information to determine if non-universal components can be placed in the Ur-IC state in light of thematic IC.

–Exploration of systems that are tied (but need not be tied) to the IC systems in question. This would include DNA replication, recombination and modification; flagella and chemotaxis; and the F-ATPases and the electron transport chain.

–Determination if there is a hierarchy of IC. That is, do some Ur-IC systems demonstrate an IC relationship with each other?

References and Recommended Reading

[1] Marczynski GT, Shapiro L.”Cell-cycle control of a cloned chromosomal origin of replication from Caulobacter crescentus.”
J Mol Biol. 1992 Aug 20;226(4):959-77.

[2] Augustin LB, Jacobson BA, Fuchs JA. “Escherichia coli Fis and DnaA proteins bind specifically to the nrd promoter region and affect expression of an nrd-lac fusion.”
J Bacteriol. 1994 Jan;176(2):378-87.

[3] Quinones et al,. 1997, Mol Micro 23, 1193-1202.

[4] Shaper, S. and Messer, W. 1995. “Interaction of the initiator protein DnaA of Escherichia coli with its DNA target.”
JBC 270: 17622-17626.

[5] Fukuoka T, Moriya S, Yoshikawa H, Ogasawara N. “Purification and characterization of an initiation protein for chromosomal replication, DnaA, in Bacillus subtilis.”
J Biochem (Tokyo). 1990 May;107(5):732-9.

[6] J.W. Schopf, 1996, “Are the oldest fossils cyanobacteria?”, in Evolution of Microbial Life, Cambridge University Press, 23-61.