Skip to content

Does ‘Dark’ Biology Have Its CHARMs?

May 3, 2011

To continue the current series of posts on ‘dark matter’ in biology, here I look at the possibility of certain hypothetical ‘dark’ replicators. This encompasses the notion of orthogonality  raised previously. In a nutshell, this asks the question: Could autonomous (self-replicating) molecular entities abide within at least some living cells and escape recognition by us until now? To approach this question meaningfully, it’s necessary to consider some background information first….

 Systems Biology and Dark Matter

 An ultimate aim of biology is to provide a coherent explanation for all biological phenomena, which will allow rational predictions of how novel agents would specifically perturb biological systems. And it is indeed at a ‘system level’ that such knowledge must be gained. In order to define the complex and interrelated networks of metabolic activity, signaling processes, and information flow which are the very stuff of biology, a systems-level description must be obtained. As noted briefly in a previous post, the new science which has this goal in mind is thus logically termed systems biology.

If the goals of systems biology are fully realized, then it could claim to be able to model life. But this can only be accomplished when sufficient experimental data has been accumulated as a necessary pre-requisite for success. This might sound obvious, but it needs to be reinforced, owing to the nature of life itself. As the product of billions of years of natural evolution, biology tends to have a ‘Rube Goldberg’ messy aspect of bits and pieces put together from an available toolkit in an ad hoc fashion. So biological systems analyses can be logically instituted, but may fail to accurately reflect reality if important pieces of information are omitted. For example, consider that until quite recently, it was believed that gene regulation was overwhelmingly mediated by protein factors operating negatively or positively at different functional levels (directly on transcription, on mRNAs, or on expressed proteins themselves). But now it is clear that while the importance of protein-based regulation is unquestionable, regulation based on RNA molecules is also extremely important.

Why does the gene regulatory tool-kit require both proteins and RNA molecules? Is there a deep reason for it, where it the most efficient evolutionary outcome possible? Or (at least to some extent), is it the result of evolutionary history and the ‘locking in’ of certain systems when alternatives would be formally possible? This is significant for a biological systems approach. If one was ignorant of RNA-based regulation, one might model an entire virtual eukaryotic biosystem based purely on protein regulators, and this might prove to be both successful and efficient. Yet in spite of this, it would not represent a real biosystem as we now understand them. And thus we return to the ‘now’ and ask whether ‘now’ is the informational end-point, or whether there are still unrecognized features of biology which are required for a realistic full systems modeling. In essence, a systems ‘map’ might in itself be a highly successful construct, but be a poor reflection of the actual ‘territory’. Of course, this in itself is a testable proposition, because the map one makes can then be used to make predictions about various aspects of the bioterritory one wishes to define. The closer the map matches the predictions, the more confident one can be that there is nothing substantial that has been left out.

With an advanced system biology, it might be possible to design more efficient biosystems which are ‘Rube Goldberg-free’ in their network construction. But here we are concerned with the totality of biology that has evolved on this planet, not future artificial ventures. And that brings us to the simple conclusion that systems biological maps will be incomplete in modeling real biological behavior as long as any ‘dark’ biomolecules or processes still exist.

The fact that data acquired regarding RNA-based regulation is relatively new might suggest that a few surprises yet await biologists. Still, as considered previously, an effective end-point for significant biological information must exist, and will probably be attained in the relatively near future. But for the time being, for the sake of argument, even if we accept that practically all the information required for a systems interpretation of biology is more or less in hand, it does not exclude the possible existence of other bio-entities existing in parallel or outside the defined system. This once again raises the issue of orthogonality, as discussed in an earlier post. Let’s now think of this further for the most intimate of possible shared environments – a membrane-bounded cell.

Intracellular Orthogonality?

 If true orthogonality for discrete organisms sharing the same physical environment is unlikely (as argued earlier), then surely this principle would be resoundingly applicable if one organism shared the cellular space of another. At first glance, this ‘sharing’ idea itself might seem contrived and highly unlikely, and yet there are ample precedents in biology for such events occurring. The best-known case in point is the origin of organelles (‘little organs’), specific subcellular structures with vital roles. Of these, mitochondria (which power eukaryotic cells through oxidative phosphorylation) and chloroplasts (the providers of photosynthesis in plants) are the most significant. Over 40 years ago, it was unexpectedly found that these organelles possessed their own small DNA genomes. The existence of these DNAs (and certain other unusual organelle properties) led to the formulation of the endosymbiont hypothesis, chiefly promoted by Lynn Margulis. This striking concept proposed that observed similarities of mitochondria and chloroplasts (and perhaps additional organelles) to generalized prokaryotic systems was a simple reflection of their evolutionary origins. These ‘endosymbiont’ organelles were postulated to have arisen from bacterial forebears in the remote past, which had invaded cells of ancestral proto-eukaryotes and eventually set up a kind of symbiotic partnership. (Here the host cell benefits from organelle biochemical activities such as photosynthesis, and the organelle gets a free ride from available nutrients and assistance with replication). During the process of mutual adaptation between host and original prokaryotic colonizer, a gradual loss of organelle genomic size and complexity occurred, with many organelle functions relegated to the host. Although initially controversial, the endosymbiont hypothesis is now widely accepted.

Of course, a great many agents which use host cells for their own replication are anything but benign. We need only think of the range of viruses, all of them essentially intracellular parasites, to be rapidly convinced of this. Even certain unusual proteins (prions) can use a host cell in order to replicate by ‘imprinting’ their aberrant conformational states onto their normal counterpart proteins (as noted in an earlier post). It is abundantly clear, then, that these agents are anything but orthogonal to their hosts once they have gained access to the host intracellular environment. And while organelles are at the opposite end of a harm-benefit spectrum, by their entwined relationship with their host cells they too are obviously and emphatically non-orthogonal replicative systems.

To return to the point made already at the beginning of this section, true biological orthogonality may be a hard call in any shared environment, and is perhaps technically impossible within the confines of a cell. But even accepting this, many degrees of minimal interaction can be envisaged short of fully orthogonal states. And the reason for considering this in the present context is to conjecture as to whether ‘dark’ intracellular replicators could exist and continue to escape detection, through having very low impact on the general activities of their hosts. In turn, the existence of quasi-independent intracellular replicators would naturally be of interest for systems biology.

For convenience, let’s give these speculative ‘dark’ intracellular replicators a name: ‘CHARMs’, for Cell-Harbored Autonomously Replicating Molecules. In the most general form of this kind of gedanken (thought) experiment, such self-replicating molecules could be anything, but in its most feasible version, the type of self-replicative molecule would be stipulated as RNA (in which case CHARM becomes Cell-Harbored Autonomous RNA Molecule). Why should this restriction be put in place? It is based on the known capabilities of specific RNA molecules to perform diverse catalyses (see a previous post), which led to the proposal of an ‘RNA World’ stage of early molecular evolution. A necessary feature of such a world of RNA-dominated biosystems would be RNA self-replication, and catalytic RNAs (ribozymes) with rudimentary RNA polymerase activity have been artificially selected.

The persistence of an RNA-based ecosystem in some specific terrestrial environment as a relic from the vastly remote RNA World has been previously conjectured, so a CHARM is simply a specialized subset of a broader speculation. Within a hierarchy of ‘biological dark matter’ topics, such RNA-CHARMs would not be at the ‘darkest’ level, but are certainly currently unknown and would be of intense interest if ever shown to exist. Armchair gedanken experiments have the great advantage of low budgets and overheads, so now we can move forward to think further about CHARMs and related issues.

CHARMs and the Meaning of Autonomy

Since ‘autonomous’ is included in the CHARM acronym, it’s useful to consider what autonomy really indicates for biological replicators. The reason for this stems from the frequent labeling of certain biological replicators as ‘autonomous’ when strictly speaking they are not. In the Table below, a series of known real intracellular replicators are listed along with hypothetical CHARMs, where several distinctions are made.

Excepting mitochondria (with a DNA genome) and proteinaceous prions, all of the remaining nucleic acid-based replicators in the above Table use RNA as their informational molecule. Accordingly, all of the RNA-based replicators require specific catalysis to replicate themselves – or in other words, specific polymerases directed by conventional RNA base complementarities. But here we see divergences as to where the information specifying the RNA-directed polymerase comes from. For conventional RNA viruses and retroviruses, the polymerase protein sequence is encoded within the respective viral genomes. But the simpler viroids make use of host-specified polymerases (See References & Details). Whether or not the infectious RNA agent specifies its own polymerase, the enzyme used for viral or viroid replication is expressed at the protein level using host translational machinery. As infectious RNA molecules which replicate at the expense of the host, viroids have been referred to as ‘autonomous’, which can be reconciled with standard dictionary definitions (usually along the lines of “acting independently or having the freedom to do so”). But clearly viroids lack complete ‘freedom of action’ in the sense that they cannot replicate without assistance from a host cell polymerase. This constraint applies to all of the replicators in the above Table, with the exception of the hypothetical ‘CHARM-2’ at the bottom. Here the autonomy of a CHARM is at the level of replication itself, in that this hypothetical RNA molecule acts as a completely self-sufficient and autonomous ribozyme polymerase, requiring only ribonucleotide triphosphate building blocks and simple cofactors (such as metal ions) present within the host cell environment.

The ‘CHARM-1’ category was added owing to a distinct possible feature of an RNA replicator, where a reverse transcribed copy is inserted into the host genome. This activity is of course exemplified by retroviruses (as noted in an earlier post), but reverse-transcribed copies of viroid-like elements (retroviroids, as in the above Table) have also been described. In eukaryotic genomes, mobile genetic elements which move themselves around (transpose) through reverse-transcribed RNA sequences (retrotransposons) are also known.  A CHARM for which a genomic copy existed would formally have two replicative possibilities: one through host RNA polymerase transcription, and the other though its in-built self-replicative potential. (If such an RNA molecule completely relied on host polymerases, then it would no longer qualify as a true CHARM by the stricter definition of ‘autonomous’). The ‘CHARM-2’ category, on the other hand, is completely independent of any host protein activity through the absence of a genomic copy. (This CHARM type could also be referred to as a Cell-Harbored Autonomous RNA Molecule, Independently Non-Genomic – or ‘CHARMING’). The ‘free RNA’ vs. ‘RNA with genomic copy’ distinction made here is somewhat analogous to the contrast between viroids, which lack genomic copies of themselves, and retroviroids, which have them. (It should be noted also that past reverse-transcription events could potentially insert defective copies of any RNA replicative element into a host genome, akin to ‘pseudogenes’).

The figure below graphically illustrates the issue of replicative autonomy for RNA viruses, viroids, and hypothetical CHARMs:

 For more details on how a CHARM might be capable of replicating itself, see References & Details.

A Low-Key CHARM Model

 Apart from organelles, all of the known real-world replicators in the above Table are pathogens, causing significant disease in their hosts. And this tends to make them visible. Even the enigmatic prions effectively flagged their own existence to biologists through their abilities to cause transmissible encephalopathies. The manifestation of some disease, some outwardly discernable abnormality, is highly probable if a blind and unconstrained replicator is unleashed within a host whose defense systems cannot immediately contain the alien threat. But what if a replicator had much more subtle effects, effectively producing no easily observable phenotype?

For example, consider a hypothetical viroid whose replication was controlled in some way, such as a self-contained feed-back regulatory loop. Unlike known viroids causing plant diseases, such a ‘tame viroid’ might persist in its host cells with little outward effect, if its copy number per cell was kept to a suitably low level. And viroids are of interest in this conjecture in other ways. Although viroids require host protein polymerases to replicate, some of them do possess intrinsic ribozyme activity for accomplishing the necessary processing of copied RNA into final mature single-stranded circular forms. This feature has led to the suggestion that at least a subset of viroids are ancient relics of the RNA World. Extending this further, a viroid with controlled copy number which could replicate itself through its own ribozyme polymerase activity converges with….a CHARM.

But could such a ‘tame’ viroid-like CHARM escape the attention of modern molecular biology for very long? Well, that depends….Obviously, if CHARMs were found only in specific tissues of only a limited number of multicellular organisms, identifying them would become proportionately harder. But this question also takes us back to the question of quasi-orthogonality once more. Using similar arguments as in a previous post, it would seem rather unlikely that a CHARM which had carried itself in association with any cellular host for a significant evolutionary period would not develop increasing functional links with the host. For example, one could propose that in the remote past, viroids did possess their own ribozyme polymerase activity, but eventually lost this function in favor of exploiting (probably more efficient) host protein polymerases. But the more the interaction with the host, the higher the ‘visibility’ of such hypothetical entities as CHARMs is likely to be. As noted, a completely parasitic CHARM producing a disease phenotype would tend to flag its own existence to outward observers, but a commensal CHARM which took only a minimal amount of a host cell’s resources would exist far more in the shadows. But on the other hand, a symbiotic CHARM which conferred benefit (in effect, a ‘molecular organelle’) would also tend to have a higher visibility.

These points aside, powerful molecular techniques for global transcriptome analyses (all transcribed RNA molecules present in a cell) are increasing being applied, and these approaches should in principle identify all RNA molecules in cells of interest. This will be extended in the next post, where we can take a critical eye to the CHARM proposal and see what’s left standing.

But at least there are still the dubious CHARMs of (biopoly)verse:

Well, could it ever be cause for alarm?

To consider such a thing as a CHARM?

But beyond mere sport

Such ventures of thought

Are beneficial, without any harm

References & Details

‘…gene regulation was overwhelmingly mediated by protein factors at different levels…’  Protein mediators of gene regulation include classic transcription factors (repressors, activators and enhancers) and a variety of co-activators and co-repressors.

‘…regulation based on RNA molecules is also extremely important.’  RNA-based gene regulation includes small interfering RNAs (siRNAs), microRNAs (miRNAs), and an increasing plethora of longer non-coding RNAs.

‘…..the endosymbiont hypothesis, chiefly promoted by Lynn Margulis.’  For a review of her original proposal, see Margulis et al. 1985. Although she is correctly credited with the development and promotion of the proposed endosymbiont origin of organelles, the actual symbiotic organelle concept is now just over a century old, first propounded in 1910 by the Russian biologist C. Mereschowsky (cited in Maynard Smith, J. & Szathmary, E. The Major Transitions in Evolution (W. H. Freeman & Co,, New York,  1995).

‘…..the endosymbiont hypothesis is now widely accepted. ‘ An important factor in this has been the application of extensive sequencing of mitochondrial and many prokaryotic genomes, allowing extensive computational phylogenetic comparisons to be made. Mitochondria from different eukaryotes show a range of sizes and complexities, with some of the larger ones being more ‘bacteria-like’. On the other hand, the process of genome reduction of organelles can go to completion, as seen with organelles of mitochondrial origin lacking genomes (mitosomes and hydrogenosomes; Embley & Martin 2006). For a considerable time, it was thought that some anaerobic single-celled eukaryotes (such as the intestinal parasite Giardia) lacked mitochondria, but more recent studies have shown that seemingly all such cases have evidence for mitochondrial genes in their nuclear material, or bear mitochondrial traces in the form of heavily modified derivative organelles (Embley & Martin 2006). This suggests that their evolutionary ancestors possessed conventional mitochondria, which have changed radically through evolution for adaptation to the life-styles of such anaerobic organisms.

‘…..with many organelle functions relegated to the host.’   Interesting precedents can be found with certain bacteria, such as the Rickettsia, which are obligate intracellular parasites. Although these organisms remain as definable bacterial species, they cannot grow in simple laboratory media, having dispensed with various biosynthetic genes in favor of exploiting host resources. Sequencing of the genome of Rickettsia prowazekii (the causative agent of human scrub typhus; see Andersson et al. 1998) was of particular interest in revealing its features reminiscent of mitochondria. In addition, R. prowazekii had more segments of non-coding DNA than other known bacteria, suggesting that these sequences were in the process of removal during the same kind of genome minimization process which the ancestors of mitochondria underwent in the remote past.

‘….an ‘RNA World’ stage of early molecular evolution.’  For a useful RNA World review, see Joyce 2002.

‘…..catalytic RNAs (ribozymes) with rudimentary RNA polymerase activity have been artificially selected.’  Initial efforts to select ribozymes with RNA polymerase function were limited to approximately 10 nucleotide additions, but very recently this has been extended to 95 bases, long enough to effect the copying of another functional ribozyme  (see Wochner et al. 2011).

‘….persistence of an RNA-based ecosystem…..has been conjectured….’  Steven Benner (who was a co-author on the ‘shadow biosphere’ paper noted in a previous post) is noted for his work in alternative nucleic acid alphabets and the RNA World, and has published extensively on this and related matters.

Viroids have been referred to as ‘autonomous’……’  For example, see Daròs et al. 2006.

‘…..reverse-transcribed copies of viroid-like elements (retroviroids) have also been described.’  See Daròs & Flores 1995.

Some Comments on Replicator Autonomy Table:

 Mitochondria:  Human mitochondrial genomes encode only 13 proteins (Calvo & Mootha 2010) and yet use their own ribosomes – but ironically the mitochondrial ribosomal proteins themselves are nuclear-encoded and transported to mitochondria by specific signal sequences. (And the same applies to proteins required for mitochondrial DNA replication). Though degenerate compared to their ancient prokaryotic forebears, organelles still have complex functions essential to their host cells. So given the co-evolution of organelles and host cells over immense time periods, it is no longer the case that they are self-replicating, even though the nuclear genes themselves which encode proteins for transport to mitochondria are likely of organelle origin.

 RNA viruses: There is considerable diversity among RNA viruses, but for the purposes of the Table all can be grouped except for RNA retroviruses.

Viroids: Viroids are single stranded circular RNA molecules of small size (typically only several hundred bases) which can infect various plant species and subvert their biosystems towards viroid replication. Such infections result in recognizable pathologies. Viroid RNA is replicated by a variants of a rolling circle mechanism and associated RNA processing to form new (+) strand circles. Among the most remarkable features of viroids is their ability to co-opt host DNA-dependent RNA polymerase to act upon their own RNA templates, since they encode no polymerase of their own (nor do they possess coding sequences for any other proteins). See Daròs et al. 2006 for a useful viroid review. The only known viroid-like entity infecting animals is the hepatitis delta virus, although this does include a single protein coding sequence (for more information, see Huang & Lo 2010).

 ‘…..(retrotransposons) are also known.’  For a review of retrotransposition and the importance of retrotransposons, see Cordaux & Batzer 2009.

Comments on Autonomy Levels Figure:

Note that the ‘rolling circle’ viroid replication depiction is simplified, since the single-stranded (+) circle must be first copied to the complementary (-) strand, which in turn templates a new (+) strand copy. This can occur by asymmetric or symmetric pathways (Daròs et al. 2006). The latter is of interest in that it involves viroid ribozyme RNA cleavages to effect processing.

Details on a model for a self-replicating CHARM:

Here the hypothetical CHARM is more specifically referred to as a self-replicating ribozyme polymerase (SRRP).

Note also: The SRRP is depicted as linear, but could be circular (as for viroids). In addition, the SRRP functions in principle could be directed by separate RNA molecules associating as a complex.

The equilibrium between folded SRRP and its denatured form is shown in favor of the folded state, which is consequently likely to be a rate-limiting step unless assisted. Unfolding and synthesis of (-) strand from (+) strand template could be coupled through the agency of the SRRP itself (i.e., one SRRP molecule binds another and partially denatures it, and then begins polymerization, which progresses through the rest of molecule acting as template). The same coupling possibility applies for templating from the duplex form, for both (+) and (-) strand templating.

Recently (as noted above), a considerable advancement in the efficiency of artificially-generared ribozyme polymerases has been reported, up to 95 bases ( Wochner et al. 2011 ). In this case the engineered ribozyme polymerase was 189 bases in length, and thus still not capable of full self-replication. But, as noted in an accompanying commentary by Michael Yarus, this goal is now a much more realistic prospect in the near future.

‘…..the suggestion that at least a subset of viroids are ancient relics of the RNA World.’  See Daròs et al. 2006.

No comments yet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s