..
..
Please Note: I do not choose the advertisements that appear with this webpage: Geocities (which provides me with free site hosting) chooses them. - Andrew Gyles
_____________________________________
T R A C I N G . H U M A N . W A N D E R I N G S
- MITOCHONDRIA and neurons
- Y CHROMOSOME
Andrew Gyles
___________________________Return to main index page
[Other page: Humans might have spoken two million years ago]
CLICK ON TITLE TO GO STRAIGHT TO ARTICLE
- Now we know: some dendrites in the brain are binary
- The mitochondrion as a flip-flop memory element in neurons
- Mitochondria as the motors of consciousness
- The evolving mitochondrion as a killer of male embryos
- The primitive mitochondrion as a fatal parasite
- Miniature apoptosis in mussels
- My hypothesis of a low mitochondrial mutation rate in humans
- Five puzzles about mitochondria
- A laboratory experiment to test the central assumption of the 'Out of Africa' theory
- A possible reason for the smallness of the human Y chromosome
ARTICLES ARE ARRANGED BELOW BY DATE OF PUBLICATION, NEWEST AT TOP
________________________________CONTENTS
Neuronal activity in discrete fields of the brain associated with the performance of specific tasks can be detected by BOLD fMRI scans (Blood Oxygen Level Dependent functional Magnetic Resonance Imaging).
The paradox revealed by these scans is that the oxygen level rises in active fields. Since neuronal activity consumes energy we should expect the oxygen level to fall. This is so because the energy is supplied in the form of ATP, and ATP is generated most efficiently by the aerobic, or oxygen-consuming, part of the long and complicated process of respiration. Aerobic respiration is done by mitochondria, which consume oxygen and generate ATP. There are hundreds or thousands of mitochondria in most cells.
The standard explanation of this paradox is that the blood vessels in an active field dilate, bringing more oxygen into the field. Little of this extra oxygen is used. It is this higher oxygen level that is detected by the BOLD fMRI scan.
A '50-50' mode in some mitochondria would reduce oxygen consumption
I suggest that there is an alternative explanation for the rise in oxygen level in an active field: some of the mitochondria in its neurons switch to a (hypothetical) '50-50' mode of action, in which one side of a mitochondrion supplies ATP to the other side, which uses this supply of chemical energy to pump positively charged protons (in the form of hydronium ions: protons bonded with water molecules) across a narrow gap to the neuronal membrane. These positive ions trigger nerve impulses. In some cases a mitochondrion might act as a 'flip-flop' memory element in neurons as well as a trigger of nerve impulses.
The oxygen consumption of mitochondria working in this '50-50' mode must be halved. One half of each mitochondrion maintains its oxygen consumption and generates ATP, but the rotary enzymes in the other half are driven in reverse, consuming the ATP and pumping out protons; this half does not consume oxygen.
If extra ATP is needed in the active field (to drive ion pumps in the neuronal membrane, for example) it must be supplied through anaerobic respiration, which does not consume oxygen. The 'standard explanation' of the raised oxygen level makes the same assumption, at least in part. It has to do this, because if most of the extra oxygen brought to the active field by the postulated dilation of blood vessels were consumed it would not show up on the BOLD fMRI scan.
Earlier articles describe possible roles of mitochondria in neurons
I have written in some detail about the possible (hypothetical) roles of mitochondria in the information-processing and signalling activities of neurons (see articles below).
One of the objections to such roles raised in scientific discussion groups was that protons were the active part of all acids, and in acting as a 'trigger' might damage the membrane of the neuron. My response to that objection is that protons drive the rotary enzymes in all mitochondria in multicellular animals while they are generating ATP. These enzymes are like molecular motors driven by a 'proton gradient' maintained by a chain of other enzymes in the mitochondrion. Obviously the protons do not damage the mitochondrion to an unsustainable degree.
The neuronal membrane maintains gradients of various ions across itself; perhaps it would be no more damaged by protons than the mitochondrial membrane. It might depend on the number of protons involved. It is known that a small number of positive (metallic) ions can trigger a nerve impulse, so a small number of protons might be able to trigger an impulse.
(I sent this article to the internet discussion group 'sci.bio.evolution' on 25 October 2001, where it was subsequently published.)
Published on this site 25 October 2001. © Andrew Gyles
________________________________CONTENTS
Now we know: some dendrites in the brain are binary
In December 2000 there was some discussion in the newsgroup "bionet.neuroscience" of my hypothesis that mitochondria might act as flip-flop memory elements in neurons (see the article "The mitochondrion as a flip-flop memory element in neurons" below).
One counter-argument ran as follows: "Why should neurons need a flip-flop memory element? There's no evidence they're digital".
Well, perhaps there is some evidence now. In the 21 September 2001 issue of "Science" Dong-Sheng Wei et al report binary behaviour of some terminal apical dendrites of pyramidal neurons, which they characterised as "all-or-none responses that were subthreshold for somatic action potentials".
They remarked that "Compartmentalized and binary behavior of parallel-connected terminal dendrites can greatly expand the computational power of a single neuron".
Of course I am aware that the paper in "Science" neither proves nor disproves my hypothesis about a possible role of mitochondria as flip-flop memory elements. However, it does show that the behaviour of some dendrites is "digital" (binary behaviour is digital behaviour), and so opens the possibility that some neurons might need and use "flip-flop memory elements".
Reference
"Compartmentalized and binary behaviour of terminal dendrites in hippocampal pyramidal neurons", Dong-Sheng Wei et al, Science, volume 293, number 5538, 21 September 2001, pages 2272-2275.
Published 25 September 2001. © Andrew Gyles
________________________________CONTENTS
In my hypotheses 'Mitochondria as the motors of consciousness' and 'The mitochondrion as a flip-flop memory element in neurons' (see below) I wrote about possible functions of mitochondria in neurons. Both hypotheses depend on my idea that the thousands of ATPsynthase/ATPase enzymes in a mitochondrion might rotate in phase in at least some situations.
Like a clock
We know that the rotating enzyme is like a clock. It rotates about 20 times a second. As it rotates it passes a positively charged ion (a proton in most organisms, or a sodium ion in one species of bacterium) through itself at every 30 degrees of rotation. So we might think of it as passing a proton every time its motor and single 'hand' come to a 'numeral' on the dial of the 'clock'; that is, at 1 o'clock, 2 o'clock, 3 o'clock and so on through 12 o'clock.
60 'big steps' a second
If it rotates 20 times a second it will pass a proton through itself 20 x 12 = 240 times a second. If it rotates in 'big steps' of 120 degrees, as some elegant work by a Japanese laboratory seems to have confirmed, it will pass four protons at each big step, and it will take 20 x 3 = 60 big steps a second.
Protons could trigger nerve impulses
When working 'in reverse' as ATPase, energised by ATP and pumping protons 'uphill' through the inner membrane of the mitochondrion, against the electrochemical gradient, the enzyme pumps protons into the cytoplasm of the cell. I argued in my hypotheses that these protons, being positively charged, could trigger a nerve impulse if some of them reached the inside of the membrane of a neuron. They could do this if the mitochondrion was inside the neuron and close to its membrane.
40,000 protons in each 'wave'?
There must be thousands of identical ATPsynthase/ATPase enzymes in each mitochondrion. Let us assume there are 10,000 (for all I know there might be 100,000). If they rotated in phase (or in phase plus or minus 120 degrees) at each 'big step' of 120 degrees they would pump a total of 10,000 x 4 = 40,000 protons into the cytoplasm. I argue that this might be more than enough positive ions to trigger a nerve impulse. (These numbers are illustrative only. I do not know what the actual numbers are. They might vary over a wide range, depending on the size of each mitochondrion.)
I posted both of these hypotheses on various scientific discussion groups. One of the criticisms made by readers was that I had not shown how they might be experimentally tested.
Response to a criticism
In response I posted the following message on the discussion groups 'talk.origins' and 'sci.bio.evolution':
________________________________________________________
A preliminary experiment to test the hypothesis that a mitochondrion might function as a flip-flop memory element in neurons
One experiment would be to search for some mitochondria that have their ATPsynthase/ATPase enzymes rotating in phase. If none can be found the hypothesis fails. Everything in it depends on the enzymes running in phase, or in phase plus or minus 120 degrees. If this happened the electrochemical gradient across the inner membrane of the mitochondrion would oscillate at about 30-70 Hz.
It does not seem an easy thing to do. The electrochemical gradient across the inner membrane of the mitochondrion is like the difference in potential across a capacitor. If the potential across a capacitor oscillated, how would you detect it except by connecting a potentiometer across its leads? There is no electric field outside a capacitor because the field extending from the positive charge on one side is cancelled by the field extending from the negative charge on the other side.
If you brought a negatively charged plate close to the positively charged side of a capacitor you would set up an electric field between the two, and so could measure the oscillating potential of the capacitor, I think. But the probe would have to be negative in relation to what?
Ideally you would insert an insulated probe with a bare tip inside the inner membrane of a mitochondrion, and then touch the outside of its inner membrane with the bare tip of another insulated probe. If the rotating enzymes were running in phase the potential across the probes would oscillate. (The outside of the inner membrane would always be positive and the inside would always be negative, but the potential difference would oscillate.)
But a mitochondrion is about the same size as a bacterium! I might be able to find an electrical engineering laboratory in Japan that would be capable of doing this small-scale work. I am confident that Japanese workers could do it. One Japanese group has already done exquisite work on the rotating enzyme.
This type of measurement was first done in 1939 in nerve fibres by Hodgkin and others. Later, glass micro-pipettes filled with a KCl solution were used as micro-electrodes.
Andrew Gyles. (Posted on the discussion groups 'talk.origins' and 'sci.bio.evolution' on 21 December 2000.)
______________________________________
In retrospect I think that I might have exaggerated the difficulty of showing that the enzymes working in reverse as ATPase rotate in phase in some or all situations. It might not be necessary to insert an insulated electrical probe into the mitochondrion.
The reason is that if the enzymes rotate in phase they will pump out protons in what I have called 'minor floods' or 'waves' about 60 times a second (60 Hz). But the electrons that counterbalance the positive charge of the protons will not be produced in 'minor floods' or 'waves' by the other enzymes in the respiratory chain; there is no reason to suppose that these electron-transferring enzymes will ever work in phase.
A detectable oscillating electric current
I therefore think that a mitochondrion with its ATPsynthase/ATPase enzymes rotating in phase as ATPase might have a small oscillating electric current that could be detected by an external electric probe. This idea could be best explained by drawings.
Detectable pressure waves
I think too that a mitochondrion with its ATPsynthase/ATPase enzymes rotating in phase would probably oscillate physically, slightly changing its shape and so generating pressure waves in the cytoplasm of the cell, at a frequency of about 60 Hz (the frequency range might be 20 - 80 hz, depending on the supply of energy-rich metabolites and oxygen, or of ATP). These waves could be detected by a tiny pressure-sensitive probe inserted into the cell and placed close the mitochondrion, but not inside the mitochondrion.
Published on this site 28 December 2000. © Andrew Gyles
________________________________CONTENTS
The mitochondrion as a flip-flop memory element in neurons
I suggested in an earlier article that if certain mitochondria in neurons worked with all of their ATPsynthase/ATPase enzymes rotating in phase or [to allow for geometric effects at the bends of cristae] in phase plus or minus 120 degrees, they would produce 'minor floods' of protons when working as ATPsynthase, which could trigger nerve impulses.
Protons are positively charged. The arrival of positive charges at the negatively charged inner surface of a neuron membrane that is ready to 'fire' will trigger a nerve impulse. The triggering positive charge need only be very small; the main strength of a nerve impulse is contributed by the subsequent increase in permeability of the membrane to sodium ions, and the inrush of that ion into the neuron.
Low mtDNA mutation rate in humans?
This article, like the earlier one ('Mitochondria as the motors of consciousness'), is speculative. The validity of both models will depend on a detailed examination of the magnitudes of the various physical quantities involved. If the models are valid they will support the notion that the mutation rate in the mtDNA of humans might be very much lower than that in apes, for reasons I outlined in the earlier article.
Triggering of 'gamma' waves of consciousness
I have suggested that a mitochondrion close to a 'critical spot' on the inner surface of the membrane of a neuron might work sometimes with its ATPsynthase/ATPase enzymes rotating in phase as ATPase, fuelled by a store of ATP in the matrix of the mitochondrion, and triggering nerve impulses in the neuron at three times the frequency of rotation of the ATPase enzymes.
I proposed that this might be the origin of the 'gamma' waves observed in certain groups of neurons. These waves have been tentatively identified by some authors as the neural correlate of 'awareness' or 'consciousness' of particular aspects of objects.
Waves with rhythm
I suggested that the triggering of such waves by a mitochondrion might continue until the mitochondrion exhausted its store of ATP, when its rotating enzymes would have to reverse their direction of rotation and work as ATPsynthase, making ATP and storing it in the matrix (the inner hollow) of the mitochondrion. This reversing of mode of operation, from ATPase to ATPsynthase, and then back to ATPase and so on would, I suggested, produce the 'rhythmic' character of the 'gamma' waves that some researchers have observed.
A more interesting mode of operation
I now propose that a mitochondrion in a particular situation inside a neuron, with one of its sides pressed close to the inside of the membrane of the neuron, might operate in a more complex manner, and one more interesting to those workers trying to find parallels between the operation of a computer and the operation of the brain.
The 'near' side and the 'far' side of the mitochondrion
I shall call the side of the mitochondrion pressed close to the inside of the membrane of the neuron the 'near' side, and the opposite side of the mitochondrion the 'far' side. I suggest that the contact between mitochondrion and neuron might be rather like the contact of two neurons at a synapse, except of course that the mitochondrion is inside the neuron (not outside it), and is smaller than a typical synapse.
There are cristae on the 'near' side of the mitochondrion and cristae on the 'far' side. These are infoldings of the inner membrane of the mitochondrion. Identical ATPsynthase/ATPase enzymes are inserted in these cristae, close to each other (I assume) in a regular arrangement a bit like the molecules in a single layer of the lattice of a crystal. There may be millions of these identical enzymes, rotating like motors in phase, in the cristae of a single mitochondrion.
'Near' side has two stable states
I propose that in particular mitochondria performing information storing and processing tasks in particular parts of a neuron the ATPsynthase/ATPase enzymes in the cristae on the 'far' side work always as ATPsynthase.
And I propose that the ATPsynthase/ATPase enzymes in the cristae on the 'near' side work sometimes in phase as ATPsynthase and sometimes in phase in reverse as ATPase, in the latter case producing 'minor floods' of protons at three times the frequency of rotation of the central asymmetric 'axle' of the enzyme. From now on I shall refer to these 'minor floods' of protons as 'waves'.
Switchable if volley phase is right
Same phase
If a volley of impulses passed through the neuron with the same frequency and phase as the waves of protons being produced on the near side of a mitochondrion with its enzymes working as ATPase the near side would 'see' high concentrations of positive charge at the peaks of the waves, and would (I propose) be 'pushed' into reverse by these high concentrations so that it worked as ATPsynthase and produced no waves of protons.
I say 'pushed' because the positive charges appearing on the inside of the membrane of the neuron would repel the protons being produced by the near side of the mitochondrion. (A nerve impulse consists of a patch of positive charge appearing to travel along the inside of the membrane of the neuron in the direction of the impulse. After it has passed, the charge on the inside of the membrane becomes negative.)
Opposite phase
If a volley of impulses passed through a neuron with the same frequency as, but opposite phase to, the waves of ATP being produced on the near side of a mitochondrion with its enymes working as ATPsynthase the near side would 'see' high concentrations of negative charge at the peaks of the waves of ATP and would be 'pulled' into reverse so that its enzymes worked as ATPase and produced waves of protons.
I say 'pulled' because the negative charges appearing on the inside of the membrane of the neuron would attract any protons in the near side of the mitochondrion and by drawing them outside it lower the concentrations of protons inside the mitochondrion.
Switchable memory element
The reader will, I think, immediately see the significance of the above two changes. In the first case the mode of operation of the enzymes in the near side of the mitochondrion is switched from a mode producing waves of positive charges to a mode producing no positive charges. In the second case the mode of operation is switched from one producing no waves of positive charges to one producing waves of positive charges.
Unswitchable if phase is wrong
If a volley of impulses passed through the neuron with the same frequency as, but opposite phase to, the waves of protons being produced on the near side of a mitochondrion with its enzymes working as ATPase the near side would 'see' low concentrations of protons (because the inside the of the neuron near the mitochondrion would be negative when each wave of protons reached its peak) and would therefore continue to work as ATPase.
If a volley of impulses passed through a neuron with the same frequency and phase as the waves of ATP being produced on the near side of a mitochondrion with its enzymes working as ATPsynthase the near side would 'see' high concentrations of positive charge at the peaks of the waves of ATP and its enzymes would continue to work as ATPsynthase.
The mitochondrion as a flip-flop
In the four cases described above the near side of the mitochondrion can be seen as analogous to a 'flip-flop', a circuit defined in United States usage as 'a bistable pair of valves or transistors, two stable states being switched by pulses ... '.
I do not push this analogy too far. Perhaps it would be better to regard such a mitochondrion as simply a binary cell, defined as 'An information storage element, which can have one or other of two stable states'. However, its current state does not have to be discovered: it is continually communicated by the mitochondrion itself.
The near side of the mitochondrion has two stable states: enzymes working as ATPase or as ATPsynthase. It can be switched between these states by volleys of nerve impulses of particular frequency and phases, but not by volleys of the 'wrong' phase relative to the current stable state.
Active memory
The near side of the mitochondrion can be seen as an active memory store, in the sense that in one of its two modes of operation it continually triggers nerve impulses, which can be communicated to other neurons, and in the other mode it triggers no nerve impulses. In the appropriate context the lack of an impulse is a signal too.
Thus it is not a passive memory store, like the patches of differing magnetic polarity on the surface of a computer hard drive, whose state has to be discovered by an actively interrogating reading head.
Short-term memory
I propose that mitochondria working as described above could be the basis of short-term memory.
Binary arithmetic
I propose that several, perhaps many, such mitochondria in the one neuron might enable it to do binary arithmetic and to process information. Perhaps it is relevant to note here that some neurons have many dendritic branches, and that a nerve impulse can travel in both directions along the membrane of a neuron.
Switching different mitochondria in the one neuron
A switching volley of nerve impulses must have a frequency three times the frequency of rotation of the enzymes in the near side of the mitochondrion. If each mitochondrion ran at its own frequency of rotation (governed, for example, by the supply and removal of metabolites in a particular situation) it could be switched only by a volley of nerve impulses of three times that frequency.
Thus incoming volleys could switch different mitochondria in the one neuron, depending on the frequency of each volley.
Andrew Gyles
(I sent this article to the internet discussion group 'bionet.molbio.evolution' on 11 December 2000, where it was subsequently published.)
Published on this site 12 December 2000. © Andrew Gyles
________________________________CONTENTS
Mitochondria as the motors of consciousness
I suggest that the reason why mitochondria have retained some protein-coding genes is that these genes code for proteins that determine, indirectly or directly, the frequencies of rotation of the ATPsynthase/ATPase enzymes in the mitochondrial cristae. I propose here that these frequencies determine the characteristic single frequencies of 'firing' of various neurons that are important to particular aspects of consciousness.
Alleles would confuse frequencies
If these genes had been transferred to the nucleus each of them would have become paired (one on each homologous chromosome). Then they would probably have mutated into functional but non-identical alleles. These alleles could be variously expressed in the mitochondria of each cell. Each cell would then contain mitochondria that rotated at different frequencies in the same circumstances. This could cause the neurons to be triggered at two or more frequencies, and so cause the loss of the aspect of consciousness for which each group of neurons was responsible.
Biparental inheritance would confuse frequencies
I propose that if mitochondrial DNA were inherited biparentally similar multi-frequency confusion would be caused in the neurons.
ATPsynthase/ATPase enzymes rotating in phase like dynamos
In certain circumstances a mitochondrion will oscillate (rather like the crystal oscillators used in electronic circuits) because 'hydraulic interference' between neighbouring F1 units of ATPsynthase/ATPase arranged in almost crystalline fashion on the cristae will force all the units to rotate synchronously and in phase. They will therefore rotate like the dynamos of a national electricity generating network, which all run synchronously and in phase.
Working in phase to avoid hydraulic interference
As an example of strict 'hydraulic interference' consider the actions of two divers reaching a completely flooded submarine and entering it, one from the stern and the other from the bow. They reach a central compartment with a door at each end. Each door can be opened inward or outward. They wish to enter this flooded compartment.
If they try to open both doors inward simultaneously they will fail. If they try to open both doors outward simultaneously they will fail. But if one diver opens his door inward and the other diver opens his door outward they will succeed. This is because water is practically incompressible and inexpandable, except for the very small compressions and expansions involved in the transmission of sound.
The divers must work together in phase so that they do not cause 'hydraulic interference'.
Turbine-dynamos driven by electrochemical gradient
Drawings and calculations elucidating the operation of the sodium ATPsynthase/ATPase of the bacterium Propionigenium modestum have recently been published (1).
This is a rotary enzyme driven by an electrochemical sodium gradient. The authors suggest that the same mechanism operates in other ATPsynthase/ATPases, including the proton-driven rotary enzyme of Escherichia coli. Much of the experimental work in this field has been done on the rotary enzymes of bacteria, the assumption being that they are similar to the rotary enzymes in the mitochondria of eukaryotes. I make the same assumption here.
Like motor-pumps when going in reverse
Rotary ATPsynthase enzymes can be driven in reverse by the energy released when ATP is split into ADP and inorganic phosphate. They then pump sodium ions [or protons, as the case may be] 'uphill' against an electrochemical gradient.
In this mode of operation the appropriate name for them is ATPase.
Three pairs of double doors
Other drawings (2) show the ATP-making (-cleaving) part of this class of enzyme to consist of three compartments. Each compartment has a catalytic site and has what looks to me like a 'double door'.
Each pair of double doors opens and closes in sequence, separated by 120 degrees of phase angle, as a central asymmetric 'axle' is rotated by the passage of positive ions through the 12-segmented rotor in the bacterial (or mitochondrial inner) membrane.
Three sites change states in sequence
According to Boyer's binding exchange mechanism, each catalytic site passes through a cycle of three different states: open, loose and tight, corresponding to an empty state, a state with bound ADP and phosphate, and a state with tightly bound ATP; at any given moment the three sites are in a different state (3).
ATPsynthase/ATPase enzymes likely to rotate in synchrony
It is reasonable to assume that the ATPsynthase molecules of a mitochondrion will rotate in synchrony because they are all identical and are all driven by the same electrochemical gradient, which depends on the concentrations of protons inside and outside the inner membrane. These concentrations in turn depend on the activity of the other respiratory enzymes in the mitochondrion.
Rotation in phase would minimise hydraulic interference
I suggest that neighbouring rotating ATPsynthase/ATPase enzymes on the cristae of a mitochondrion will fall into a phase relationship that minimises 'hydraulic interference' between neighbours as their opposed 'double doors' open and shut. There might be millions of these enzymes in a single mitochondrion.
Frequency determined by supply of metabolites
Thus all of the identical rotary enzymes at the end of the respiratory chain will rotate in phase at a frequency ultimately determined by the supply of incoming metabolites and the metabolic activity of all of the 'earlier' enzymes in the respiratory chain. I suggest that the full range of frequencies of rotation of the ATPsynthase/ATPase might turn out to be about 6 - 24 Hz.
Typical frequencies of rotation
The results of the calculations in (1) were that when the sodium-driven rotary enzyme operated under a typical physiological electrochemical gradient it rotated at about 15 - 20 Hz. It formed 3 molecules of ATP per revolution. Twelve sodium ions passed through the rotary Fo part of the enzyme (the 'motor') for each revolution, corresponding to an average of 4 sodium ions per ATP molecule formed. The authors remarked that the same operating principle could drive the proton-driven ATPsynthases, with slight changes to the design of the Fo motor.
Other workers observing an isolated F1-ATPase unit by video microscopy (4) have shown that it rotates in steps of 120 degrees.
I conclude from these papers that it is likely that the proton-driven ATPsynthase enzymes in the cristae of mitochondria rotate at about 15 - 20 Hz (the precise rate depending on the electrochemical gradient driving them). It is possible that when their movement is looked at in fine detail they will be seen to advance in 12 small steps of 30 degrees per rotation (as shown in a figure of reference 1), at each step releasing one proton at the bottom of the gradient. However, the 30-degree steps shown in (1) are derived from theoretical calculation; they have not, as far as I am aware, been observed.
Rotation in phase in 'big steps' of 120 degrees
I assume that when the rotation of the ATPsynthase/ATPase enzyme is looked at more broadly it is seen to advance in three big steps of 120 degrees per rotation (as observed in reference 4), at each big step releasing one molecule of ATP and four protons at the bottom of the gradient. I propose that all of these enzymes in a mitochondrion rotate in phase, or in phase plus or minus 120 degrees; that is, they all take a 120-degree step at precisely the same time.
Frequency of 'big steps' same as gamma waves of consciousness
The rate of the big steps will therefore be about 45 - 60 Hz. There is an important coincidence here: this rate is about in the middle of the 'gamma' range of oscillation (20 - 70 Hz) of certain neurons that seem to play important roles in various forms of perception, awareness, consciousness, and motor control. The precision of the timing of the cycling of these oscillations has been observed to be about a millisecond.
Some workers have reported 'rhythmic' activity of neurons oscillating at gamma frequencies; I assume that they mean regular starting and stopping of the oscillations (5, 6, 7).
As I suggested above, the full range of frequencies of rotation might turn out to be 6 - 24 Hz (at least while the enzyme is operating as ATPase); this would make the range of frequencies of the 'big steps' 18 - 72 Hz and thus cover the full range of gamma oscillations.
Enzymes in reverse pump 4 protons uphill at each big step
If the ATPsynthase/ATPase enzymes are driven in reverse as ATPase, utilising the energy released by the splitting of ATP, the same frequency relationships will hold, I assume.
The rotation rate will be about 15 - 20 Hz (perhaps 6 - 24 Hz). One ATP molecule will be split for each big step of 120 degrees of rotation. Each enzyme will take 15 - 60 (perhaps 18 - 72) big steps of rotation per second, precisely in phase with its neighbours, and in each big step 4 protons will be pumped up the electrochemical gradient to the outside of the inner membrane of the mitochondrion.
How might a nerve impulse be triggered?
How might the (hypothetical) oscillations in a mitochondrion caused by the ATPsynthase enzymes rotating synchronously in phase in the cristae trigger oscillations of the same frequency in the 'firings' of a neuron?
Possibility one: working as ATPsynthase
One obvious possibility is that the mitochondrion will produce minor 'floods' of ATP molecules at frequencies in the range of 45 - 60 Hz. The neuron uses ATP to drive the enzymes in its membrane that repolarise the membrane after each 'firing'. If just enough ATP molecules arrive at the membrane in minor 'floods' with a frequency of (say) 45 Hz, the membrane could be repolarised at that frequency and so could not fire at any higher frequency. However, it could fire at a lower frequency. And in any case, because ATP is made inside the inner membrane of the mitochondrion, and has to be carried outside by the ATP/ADP translocators in that membrane, it is possible that an oscillation in its supply inside the mitochondrion will be almost completely 'smoothed out' or 'averaged' by the time the ATP arrives outside.
Possibility two: working as ATPsynthase
Another possibility is that the ATPsynthase enzymes rotating in phase will continually lower the concentration of protons outside the inner membrane of the mitochondrion at a frequency in the range of 45 - 60 Hz. When operating as ATPsynthase they are driven by the electrochemical gradient created by a high concentration of protons outside the inner membrane and a low concentration of protons inside it. (Other respiratory enzymes in the mitochondrion keep building up the concentration of protons outside the inner membrane.) Inside the membrane there may be a corresponding oscillating change in the concentration of protons as protons are released in minor 'floods' from the ATPsynthase enzymes.
'Minor floods' of protons as triggers
If a sufficient number of excess protons passed through the outer membrane of the mitochondrion (which has VDAC pores that make it freely permeable to ions and most metabolites) (8) in minor 'floods' at frequencies in the range of 45 - 60 Hz and reached a critical spot on the membrane of the neuron they could depolarise the membrane at that spot and so trigger impulses at the same frequency.
A neuron membrane ready to 'fire'
The membrane of a neuron that is ready to fire is polarised by negative charges inside it and positive charges outside it. Protons, which are positively charged, arriving at the inner surface would depolarise the membrane and so cause the neuron to fire.
Some doubts
The question here is whether the ATPsynthase enzymes in a mitochondrion inside a neuron would allow minor floods of excess protons to leave the mitochondrion and cross a (small) gap to the inside of the membrane of the neuron. ATPsynthase might pass protons so quickly from outside the inner membrane of the mitochondrion to inside that no excess protons could pass from the mitochondrion to the membrane of the neuron. And it is possible that the positive charge of the protons outside the inner membrane of the mitochondrion is balanced, partly or wholly, by the negative charge of anions inside the inner membrane, so that the protons on the outside are 'tied' by the attractive force between them and the anions. (That is to say, the inner membrane of the mitochondrion might be partly or completely polarised when ATPsynthase/ATPase is working as the synthase. I have no information on this.)
Possibility three: working as ATPase and pumping protons in waves
Because of the doubts outlined above I think it likely that, if mitochondria are indeed the motors of consciousness, they trigger the rhythmic firings of certain neurons at the gamma frequencies only when the rotary enzymes are working as ATPases, fuelled by ATP and pumping protons from inside the inner membrane of the mitochondrion to outside it.
'Minor floods' or waves of protons
I propose that when working thus they can produce 'minor floods' or waves of excess protons that can quickly cross a small gap from the mitochondrion to the inside of the membrane of a neuron. The 'floods' of protons arrive at a gamma frequency and trigger firings of the neuron at the same gamma frequency.
As I explained above, the membrane of a neuron that is ready to fire is polarised by negative charges inside it and positive charges outside it. Protons, which are positively charged, arriving at the inner surface would depolarise the membrane and so cause the neuron to fire.
Reversing enzymes give gamma waves their rhythm
When the mitochondrion exhausts (or nearly exhausts) the supply of ATP inside it the ATPase has to stop using ATP as a fuel; it then rotates in the reverse direction and is driven as ATPsynthase by the flow of protons from outside the inner membrane to inside.
Thus the neuron will be triggered at a gamma frequency until the mitochondrion runs out of ATP. Then the triggering will stop for a while, while the mitochondrion 'recharges' its supply of ATP. This explains the rhythmic pattern observed in the gamma oscillations.
Natural selection for a low mitochondrial DNA mutation rate
If mitochondria are the motors of consciousness because they can drive certain neurons each at a single gamma frequency characteristic of that neuron at any one time (or of a group of neurons of which it is a member) it is possible that there has been natural selection for a lower mitochondrial mutation rate in humans than in other animals. This is so because higher consciousness (especially the ability to use language) gives humans many advantages in the struggle to survive and reproduce.
Mutations (either maternally inherited or somatic) in mitochondrial DNA could produce populations of mitochondria that oscillated at different frequencies in certain single neurons important in the creation of an aspect of the conscious state. The driving of a single neuron at two or more different gamma frequencies would presumably destroy the aspect of consciousness that its firing would normally help to create.
Information systems in single-celled organisms
I suggest that rotary enzymes might play important roles in the simple information systems of bacteria and single-celled eukaryotes. The slime mould that extends itself through the shortest path in a maze, for example, might depend on a sonar system driven by the oscillations of mitochondria.)
References
1) Dimroth P et al (1999), Proceedings National Academy of Sciences 96(9): 4924-4929.
2) Scientific American, January 1998, p 9.
3) Saraste M (review) (1999) Science 283:1488-1493.
4) Noji H (essay with drawing) (1998) Science 282:1844, 1845.
5) Maldonado P E et al (2000) Cereb Cortex 10(11): 1117-1131.
6) Palva J M et al (2000) J Neurosci 20(3): 1170-1178.
7) Mima T et al (1999) Neurosci Lett 275(2): 77-80.
8) Green D R et al (1998) Science 281:1309-1312.
Andrew Gyles
(I sent this article to the internet discussion group 'bionet.molbio.evolution' on 30 November 2000, where it was subsequently published.)
Published on this site 1 December 2000. © Andrew Gyles
________________________________CONTENTS
The evolving mitochondrion as a killer of male embryos
In any species of animal in which mitochondria are inherited only from the mother a conflict exists between the selfish interest of the mtDNA and the interest of the species.
The selfish interest of the mitochondrion and its mtDNA is to be transmitted by every mature individual of the species to the next generation. But in the case of uniparental maternal inheritance of mtDNA half of the individuals, the males, do not transmit their mtDNA to the next generation. And in the more complex animals the mothers expend much energy and time in bearing each offspring.
In this situation if a mitochondrion in a male embryo could think, and could detect the sex of the embryo, it would realise that the best chance of ensuring the transmission of its mtDNA to the next generation would be to kill the embryo. It has zero chance of being transmitted to the next generation if it is in a male embryo. If it kills the male embryo the mother will soon have another chance to conceive, and there is a 50 per cent chance that her next offspring will be a female, which will transmit the same mtDNA to the next generation in the fullness of time. This is true because the mtDNA in the mitochondrion in the male embryo is identical to the mtDNA in the mother.
Of course mitochondria cannot think, but natural selection acting on random mutations of mtDNA (like throws of dice on the board of the conditions of existence) could produce the same effect. The mtDNA in an animal species in which inheritance of mtDNA is uniparental and maternal will evolve so that the mitochondria can detect the sex of an embryo and, if it be male, kill it. The killing mechanism would, I suggest, have been similar to the "miniature apoptosis" that destroys male-line (M type) mitochondria in female embryos in mussels. (I wrote about this hypothetical miniature apoptosis in mussels in another article.)
In the present case the maternally inherited (F type) mitochondria would be destroyed, leaving the cell with no mitochondria. The embryo would die.
The result of this evolution would be that most of the offspring born in each generation would be female. But this would not be good for the species as a whole. In the long run natural selection would favour those populations of individuals in which the mtDNA genes coding for proteins that detected and killed male embryos had been "consficated" by the nucleus of the cell and brought under the control of the nucleus. I suggest that this is the reason why humans, for example, have only 13 protein-coding genes left in the mtDNA, all of them coding for respiratory enzymes. The mitochondria have been disarmed.
In my hypothesis on miniature apoptosis in mussels I suggested that female-line (M type) mitochondria might have receptors in their outer membrane for a protein that was a receptor for the female-steroid molecule, and that in the abscence of the female-steroid molecule this protein fitted specifically into the receptor in the outer membrane and permeabilised it, thus destroying the mitochondrion. Such a mechanism might help the species by causing the death of any mussel embryo, whether male of female, that did not quickly produce a typical level of female-steroid molecules.
How could a mechanism like this in the ancestors of species that inherit all of their mtDNA maternally be modified by evolution to bring about the destruction of male embryos by miniature apoptosis? I think that the protein that was a receptor for the female-steroid molecule would have to evolve so that when it formed a specific complex with a male-steroid molecule it fitted specifically into the receptor in the outer membrane of the mitochondrion and permeabilised the membrane.
Andrew Gyles
(I sent this article to the internet discussion group 'bionet.molbio.evolution' on 2 November 2000, where it was subsequently published.)
Published on this site 3 November 2000. © Andrew Gyles
________________________________CONTENTS
A possible reason for the smallness of the human Y chromosome
In humans the gamete contributed by the father determines the sex of the child. If the gamete has a Y chromosome the child is male. If the gamete has an X chromosome the child is a female.
It would be in the selfish interest of the Y chromosome to get itself passed to much more than 50 per cent of the next generation. It is perhaps possible that more than once in the course of evolution of humans (and their non-human ancestors) mutations in the Y chromosome have resulted in genes that discriminate against the production of X- bearing gametes in spermatogenesis or sperm maturation. The most recent of such events might (for instance) have occurred in Africa 60,000 years ago.
Such mutations would give the Y chromosome a tremendous evolutionary advantage. It might ensure that (for instance) four out of six of a man's offspring were males. This mutated chromosome could spread quickly through a widespread established population, generation by generation (assuming that a short-displacement wavelike movement of people, lasting a long time, was always possible).
However, no other chromosomes would accompany it far on its journey. It would be a bit like the gene for 'hornlessness', which can be introduced into a herd of horned black Angus cattle by a cross with a hornless red Shorthorn bull and, under the artificial selection of the cattle-breeder, be spread throughout the herd. In about 20 years the herd consists of hornless black Angus cattle. It would be incorrect for a stranger to assume that the black cattle he or she sees in the field were of a breed that had been hornless for hundreds of years.
Under natural selection a human population in which the males had the 'anti-X gamete' Y chromosome would suffer the disadvantages of a surplus of males and a scarcity of females. In the long run natural selection would favour subsequent mutants of the 'anti-X gamete' Y chromosome in which those mutant genes that discriminated against the production of X-bearing gametes in spermatogenesis or sperm maturation were deleted or rendered inactive.
The result after many millenia of evolution of humans (and their evolutionary precursors) would be a human Y chromosome from which many genes had been deleted and in which many of the remaining DNA sequences were permanently inactivated. It would also be a Y chromosome that could give a misleading picture of the chronology and routes of human migration from Africa.
Andrew Gyles
(I sent this article to the internet discussion group 'bionet.molbio.evolution' on 1 November 2000, where it was subsequently published.)
Published on this site 3 November 2000. © Andrew Gyles
________________________________CONTENTS
Miniature apoptosis in mussels
Mussels have two types of mitochondria and mtDNA: one that is transmitted from the mother to daughters and sons (the F type), and one that is transmitted from the father to daughters and sons (the M type).
The female and male embryos receive M mtDNA through the sperm, but within 24 hours it is eliminated or drastically reduced in female embryos but is maintained in male embryos (1).
I suggest that M mitochondria have in their outer membrane molecules that specifically fit a protein that acts as a male-steroid receptor. When this protein is present in the cytoplasm of the cell in the absence of male-steroid molecules it binds to the outer membrane of M mitochondria and permeabilises it, thus destroying the M mitochondria. I call this action 'miniature apoptosis' because the tiny organelles are destroyed, but not their host cell. When this protein forms a complex with a male-steroid molecule its conformation changes, so that it cannot permeabilise the outer membrane of M mitochondria.
Thus M mitochondria are destroyed in female embryos, whose cells contain no, or few, male-steroid molecules. But M mitochondria are not destroyed in male embryos because their cells contain many male-steroid molecules.
It is possible that F mitochondria have in their outer membrane molecules that specifically fit a protein that acts as a female-steroid receptor. When not complexed with a female-steroid molecule this protein can permeabilise the outer membrane of F mitochondria. When complexed with a female-steroid molecule the protein cannot permeabilise the F mitochondrial outer membrane. If the cells of both female embryos and male embryos contain many female-steroid molecules the F mitochondria will be maintained in both sexes.
This hypothesis is, I think, given some support by findings in a paper published last August. In human cells a steroid receptor called TR3, normally present in the nucleus, translocates to mitochondria in response to apoptotic stimuli and permeabilises the mitochondrial membranes, releasing cytochrome c and triggering a series of events that cause apoptotic cell death. TR3 is called an 'orphan' receptor because the steroid with which it is assumed to form a specific complex has not yet been identified (2).
Humans inherit mitochondria and mtDNA only from their mothers. I suggest that the evolutionary 'ancestor' of TR3 was like the (hypothetical) female-steroid receptor in mussels that permeabilises the outer membrane of F mitochondria in the absence of female-steroid molecules. I suggest that TR3 has evolved in such a way that it can no longer form a specific complex with a female-steroid molecule, and is therefore able to permeabilise mitochondria and trigger apoptosis in all human cells, whether male cells or female cells. It is a permanent 'orphan'.
1. Sutherland et al, Genetics 148: 341-347 (January, 1998).
2. Li et al, Science 289: 1159-1164 (18 August, 2000); and in the same issue the perspective by Brenner and Kroemer, pp 1150 and 1151.
(I thank group members who provided information and references on mitochondrial inheritance.)
Andrew Gyles
(I sent this article to the internet discussion group 'bionet.molbio.evolution' on 30 October 2000, where it was subsequently published.)
Published on this site 31 October 2000. © Andrew Gyles
________________________________CONTENTS
The primitive mitochondrion as a fatal parasite
The primitive mitochondrion might have been a parasite of primitive eucaryotic cells that preserved a free-living stage in its life cycle, gaining its freedom by lysing its host.
Such an association would not have been disastrous for the host species if it were able to multiply more quickly with the parasite than without it, and if the parasite did not lyse a cell of its host too frequently. The great gain in metabolic efficiency conferred by the oxidative respiration of the primitive mitochondrion would have helped the host cell to grow more quickly and multiply more quickly, and therefore a single-celled host species could survive the occasional lysing of some of its cells.
Nonetheless, the host would gain a further advantage if it could stop the primitive mitochondrion from lysing the cell of its host. As the primitive single-celled eucaryotic cell evolved into separate species each species might have developed its own ways of 'disarming' its parasite. One way would have been to 'confiscate' those genes of the parasite that were involved in the lysing of the host and place them in the nucleus under the control of the host. Another way might have been for the host cell to interfere in the expression of the genes that remained functional in the mitochondrion by sending host-derived control proteins or RNA into the mitochondrion.
The evolution of multicellular species could not have proceeded far if occasionally some of their cells were lysed at random. Such random lysing would disrupt the organisation upon which multicellular species depend for their success. Therefore we might expect to see in these organisms a more stringent control of lysing by mitochondria than in single-celled species.
However, controlled lysing of particular host cells might have been of great advantage in the evolution of multicellular species. Such an advantage might have been achieved if the host gained control of the lysing activity of its mitochondria and applied it in a programmed way to achieve 'programmed cell death', or apoptosis. And indeed we know that mitochondria are the focus of many of the drastic processes involved in apoptosis.
Thus the simple assumption that the primitive mitochondrion was a parasite that occasionally lysed the cell of its host can explain why some mitochondrial genes have been transferred to the nucleus in many species of eucaryotic cell, and why mitochondria are the focus of apoptotic activity in various multicellular species.
(Note: I have used the terms lysing and lyse in a general sense. I have used words such as disarming, confiscate, interfere, and so on as a concise way of expressing the results under natural selection of a few trillion dice throws on the board of the conditions of existence.)
(I sent this article to the internet discussion group 'bionet.molbio.evolution' on 28 September 2000, where it was subsequently published.)
Published on this site 29 September 2000. © Andrew Gyles
________________________________CONTENTS
My hypothesis of a low mitochondrial mutation rate in humans
There is a theoretical possibility that a mitochondrion could eliminate those mutations that cause a distortion of the double-stranded mtDNA, because such mutations are physically detectable by a set of enzymes.
The enzymes could then destroy the entire double-stranded copy of the circular mtDNA genome containing the distorted mtDNA, or alternatively tag it for 'export' and cause it to be ejected from the mitochondrion. (The ejecting of genetic material through a conjugation tube is a known bacterial behaviour. Mitochondria are thought to have a bacterial ancestor.)
The destroyed or ejected copy of the mtDNA genome could be replaced, sooner or later, by the replication of an unmutated copy in the same mitochondrion.
It is interesting to note that enzymes in the nucleus try to do the same thing. But of course, having detected a chromosome distorted by a mutation they cannot destroy the chromosome. If they did that they would not have an unmutated copy to replicate and so make good the loss. The homologous chromosome is not an identical copy. So the best they can do is to 're-pair' the distorted part of the double-stranded nuclear DNA. This is a chancey business because they cannot 'know' which base or bases to leave in and which to cut out. But it is better than doing nothing.
Oxygen free radicals produced by the respiratory chain in mitochondria can damage mtDNA and thus increase the mutation rate. It is conceivable that an abnormally high mtDNA mutation rate in somatic cells is an indication that this hypothetical system for eliminating mutations that cause distortions in double-stranded mtDNA has failed in those cells.
It is worth noting that if such a system is at work in the mitochondria of female germline cells it might have evolved to different degrees in different species of animal.
For example, the development of higher intelligence and memory and the acquisition of language in humans might have depended on a simultaneous evolution of a lower mtDNA mutation rate in human brain cells in particular. However, the lower rate might have been achieved generally, in all cells. In that case it is possible that the mtDNA mutation rate in human female germline cells is much lower than has been assumed in studies of human evolution based on the 'mtDNA sequence divergence rate'. As far as I know this rate has never been objectively measured.
Some indication of whether mitochondria do use the hypothetical system outlined above could perhaps be gained on cell cultures in vitro, using X-irradiation as a mutagenic agent to accelerate the mutation rate.
(I published this summary on the 'message board' of the discussion group 'talk.origins' on 10 September 2000.)
Published on this site 10 September 2000. © Andrew Gyles
________________________________CONTENTS
Five puzzles about mitochondria
The first puzzle is: Why does each mitochondrion have several copies of its chromosome? Each copy resembles a bacterial chromosome: it is circular and has no histones. But a bacterium has only one copy of its chromosome.
I think that I have already solved this puzzle (see my letter to the Editor of 'Nature', sent on 2 August 2000, below). The mitochondrion has several copies of its chromosome so that it can destroy (or, alternatively, as I realised a day or two ago, eject from the mitochondrion) any copy in which a mutation has caused a physically detectable distortion, and then replace the destroyed or ejected copy by replicating an unmutated copy.
As I remarked in my letter to the Editor of 'New Scientist' sent on 14 July 2000 (see below), conditions inside a mitochondrion are thought to be hostile to DNA; oxygen free radicals produced by the respiratory chain can damage DNA, causing mutations. Though mitochondria are thought to be descended from a bacterium (or from a common ancestor of bacteria, mitochondria and chloroplasts) the conditions inside a mitochondrion are likely to be much more hostile to DNA than those inside a bacterium. The mitochondrion is a highly specialised, hard-driven oxidiser of metabolites. Therefore, as I have suggested, it needs several copies of its chromosome.
The second puzzle is: Why does the mitochondrial chromosome of most species of animal contain so few genes? For example, in humans it has 37 genes, of which only 13 code for proteins. The most 'primitive' mitochondrion yet discovered, in the flagellate Reclinomonas americana, has 92 genes (1). Of these, 27 code for ribosomal proteins, one codes for an elongation factor and four code for components of RNA polymerase. None of these 32 genes has been retained in the mitochondrion of humans. Many of them have been transferred to the nucleus. Why?
One answer must be that it is safer to keep the genes in the 'mild' environment of the nucleus than in the 'hostile' environment of the mitochondrion (see the argument under 'first puzzle' above).
However, of the 24 genes in the Reclinomonas americana mitochondrion that code for proteins involved in oxidative respiration the human mitochondrion has retained 13. Why?
One author has written that the mammalian mitochondrial genome has been reduced in size in order to increase its replication rate. He wrote that many essential mitochondrial genes had been transferred to the nucleus of the cell, and 'The transfer of mtDNA sequences to the nucleus continues to this day' (2). I think that this is an over-simplification.
The real question, in my view, is not 'Why have so many mitochondrial genes been transferred to the nucleus?' but 'Why have not all of the mitochondrial genes been transferred to the nucleus, and why were those coding for respiratory proteins nearly always retained ?'
I argue that there has been plenty of time for all of the protein-coding genes of the mitochondrion to be transferred to the nucleus, but the cells that allowed this to happen have not been successful in the struggle for survival. The reason why remains the greatest mystery in the five puzzles.
The third puzzle is: Why do the mitochondria in a cell occasionally fuse together?
This behaviour is a bit like the way a bacterium sometimes donates a copy of some of its DNA to another bacterium through a conjugation tube. Has mitochondrial DNA been observed passing from one mitochondrion to another when the mitochondria are in the fused state?
The fourth puzzle is: Why does only the female in multicellular animal species transmit mitochondrial DNA to the next generation?
One answer is that the female's ovum is much bigger than the male's sperm and can therefore contain vastly more copies of the mitochondrial chromosome.
Another is that the sperm has to work very hard in its voyage to the ovum and during this great physical effort many of the copies of its mitochondrial chromosome are damaged.
I doubt that either of these is the main answer to the puzzle. I think it more likely that multicellular animals that allowed both sexes to transmit their mitochondria to the next generation did not succeed in the struggle for survival: only one sex could be allowed to do it, and the reasons given in the above two answers made it certain that the sex that was allowed to do it was the female. The reason why only one sex should be allowed to transmit copies of the mitochondrial chromosome to the next generation remains a mystery.
The fifth puzzle is: Why do mitochondria play an important part in apoptosis, programmed cell death? This is sometimes called cell suicide.
As I remarked above, I think that I have solved the first puzzle. I have spent some time thinking about the remaining four puzzles and consider that they are susceptible of solution, though not without imagination.
References
1) Palmer, Jeffrey D., Nature, 387, 454-455 (1997).
2) Wallace, Douglas C., Science, 283, 1482-1487 (1999).
Published 20 August 2000. © Andrew Gyles
________________________________CONTENTS
A laboratory experiment to test the central assumption of the 'Out of Africa' theory
I suggest that female chimpanzee germline cells be grown in a glass dish and female human germline cells be grown in another glass dish. (If it is not possible to grow germline cells somatic cells will have to be used instead.)
The cells in both dishes should then be treated with equal doses of mutation-inducing radiation, such as X-rays, for a calculated period of time. This period should be long enough for the radiation to cause mutations in a few of the mitochondrial chromosomes of some of the cells but not long enough for it to cause mutations in the nuclear chromosomes of many of the cells. This should be possible because there are many more mitochondrial chromosomes than nuclear chromosomes in most cells. (For example, some cells contain thousands of mitochondria, and each mitochondrion contains several copies of the mitochondrial chromosome. I do not know typical figures for female chimpanzee and female human germline cells.)
Many experiments have studied the mutagenic effects of irradiation with various doses of X-rays, so the choice of appropriate doses can be guided by a great reserve of experience. Alternatively, specialists in the experimental use of mutagenic agents might be able to suggest more suitable mutagens for this experiment.
After the irradiation has been stopped the cultured cells in each dish should be allowed time to repair or destroy mutated mitochondrial chromosomes. Then their mitochondrial DNA (mtDNA) should be searched for permanently mutated sequences and the net rates of mutation of mtDNA in the chimpanzee cells and the humans cells should be calculated and compared.
The central assumption of the 'Out of Africa' theory is that the sequence divergence rate in human mtDNA (determined by the mtDNA mutation rate) is about the same as that in chimpanzees. But I suggest that humans correct mutations to their mtDNA much more than do chimpanzees (see note below). If this experiment shows that the net sequence divergence rate in humans is about a tenth of that in chimpanzees it will remove the logical support for the 'Out of Africa' theory that this assumption has always provided.
The experiment might be a laborious and difficult one. It is of course an artificial one that speeds up the process of mutation millions of times. There might be questions about the choice of cells to be used, the correct mutagenic treatments and the interpretation of the results. However, until such an experiment is done it remains true that the central assumption of the 'Out of Africa' theory has not been put to the acid test in the laboratory.
(Note: A typical mammalian cell, whether somatic or in the female germline, contains hundreds or thousands of copies of the single mitochondrial chromosome. But no normal mammalian cell contains more than one copy of each nuclear chromosome except during replication: nuclear chromosomes are paired in somatic cells, but the members of the pairs are not identical. The simplest way for a mitochondrion to correct a mutation to its mtDNA would be for it detect any mitochondrial chromosome containing a mispaired region and destroy the entire chromosome. There are several chromosomes in each mitochondrion, and a new chromosome could be produced by the replication of an old one when required. Let me emphasise that this is not possible in the chromosomes of the nucleus.
Another way for a mitochondrion to correct a mutation in its mtDNA would be for it to force its several chromosomes to swap strands, then repair the mispaired region that would be revealed as new pairs of DNA strands formed. There would always be a risk that the mutation might be spread by its being taken as the correct template to which to fit nucleotide bases during enzymatic repairing of the mismatch, but this would be numerically unlikely. It would be rendered virtually impossible if all of the mitochondria in a cell fused for a time and forced their chromosomes to swap DNA strands widely and then form new strand pairs. This would reveal mutated strands because they could not form a perfect pair with a normal strand. The cell could then either destroy the chromosomes containing mispaired bases, or repair the mismatched bases [the latter process being less than perfect because the repairing enzymes can hardly be expected to 'guess' correctly every time which is the correct base or sequence of bases to use as the template in performing the repairing. However, if the cell repeatedly forced the mitochondrial chromosomes to swap strands while the mitochondria were fused, and successively repaired mispaired strands, mutated strands would eventually be corrected].
The mitochondria in a cell do fuse from time to time. I suggest that they do indeed destroy or repair mutated mtDNA strands while fused. Fusing is perhaps a bacterial behaviour inherited from their bacterial ancestors, which swapped or donated sections of DNA through conjugation tubes. Nature is a great opportunist, and might have adapted the conjugating behaviour of the bacterial ancestor, in which one bacterium could donate a mutant gene to another bacterium, to similar behaviour in fused mitochondria, in which mutant mitochondrial DNA was not donated but revealed or exposed and got rid of.
This [hypothetical] process of correcting mutations to mtDNA while the mitochondria of a cell are fused might interfere with the vital energy-converting function of the mitochondria. I suggest that an animal would evolve so that its cells did it only as often as necessary. And I suggest that something in the evolution of humans, perhaps the acquisition of language and abstract thought, and the great need that this created for perfect functioning of the brain cells over a period of many years, caused humans to correct mutations to their mtDNA much more effectively and rigorously than chimpanzees, and that this is done in all human cells, not just brain cells. Alternatively, the mitochondria might be involved in brain function in humans in a more complicated way than in other animals, and this might cause them to behave in the way I have outlined above. I shall write about this possibility in another article.)
Published 19 June 2000. © Andrew Gyles
________________________________CONTENTS