BIOLOGY

ORIGIN OF INTELLIGENT LIFE :PART 1

What we see is a solution to a computational problem,our brains compute the most likely causes from the photon absorptions within our eyes(Hermann Helmholtz)

Unicellular intelligence

The first eukaryotic cells could phagocyose food including bacteria. This was due to their novel membrane system. The membrane could deform and surround and engulf the food particle and later digest it by secreting enzymes  in the vesicle. With the help of a cytoskeleton which was self organizing and formed microtubules towards advancing side while dissolving them on rear side,amoeboid locomotion was possible. Other types of locomotion depended on rows of cilia or a flagellum. The flagellum is a complicated system and has a rotating motor at the base of the flagellum driven by ATP. It exists in bacteria also and enables these bacteria to move towards food.

    With locomotion there is a need to coordinate movement towards a food after its identification. This is based on chemoattraction and repulsion through a chemical gradient. This phenomena is also seen in bacteria which can move with a flagellum. .However the bacteria does not respond to level of concentration but only to a gradient. This means at high concentration the  bacteria stops moving and adapts to the high levels and instead does random tumbling motion. This allows the organism to detect gradients and move towards  them when present. A protein circuit leads to this exact adaptation.The receptor when combined with a chemoattractant inhibits phosphorylation of a protein  phosphorylase and this inhibits flagella tumbling and leads to movement towards the gradient. Over a period of time the receptor paradoxically increases methylation of phosphorylase leading to increased activity of the enzyme and increased tumbling.This happens because the receptors inhibits the enzyme which degrades the methylating enzyme of the phosphorylase.

Primitive nerve networks

With rise of multicellular organisms movements had to be coordinated in the cellular conglomeration. The first animals were sessile.These were the phyllum Porifera or sponges . Sponges have a mass of cells organized into pores and channels through which water with food  particles flow in and out, The food particles while passing through are phagocytosed by the sponge cells. There is specialization in sponge cells for various purposes. For locomotion they have specialized myocytes. These cells contains myofibrils and can contract. They are present around pore openings. They respond to chemical signals in water and regulate pore opening to increase or decrease flow of water .

  Further differentiation into nerve cells occur in phyllum coelentrate which have true locomotion.The typical example is hydra.The hydra has a number of tentacles which it can use for locomotion by a tumbling motion.It can also use the tentacles to grasp food and bring it to the oral cavity.The hydra contains neuronal networks. Nerve nets contain sensory neurons which can respond to chemicals,light or touch.It also contains motor neurons which control individual  groups of muscles based on direction of movement. Inter neurons coordinate between sensory neurons and motor neurons or different networks.Another organism with multiple networks is Caenorhabditis elegans belonging to phyllum platyhelminthis or flatworm. For example locomotion in C. elegans is carried out by a forward neural network and reversal by two redundant  networks. Each network contains sensory motor and interneuronsns. Touching the worm produces a reversal movement followed by a detour and forward crawling. This sequence is brought about by intial inhibition of forward network and simultaneous activation of reversal network, subsequently the forward network is again activated. 

Learning in innate nerve networks

Innate networks can learn  a changing environment to a certain extent. One mechanism is habituation. If a C elegans is repeatedly tapped it reduces the frequency and intensity of tap withdrawal. Habituation acts to change synaptic transmission and produces decreased response to a repeated routine stimuli.

Other mechanisms to change network coordination has been found in higher organisms like lobster belonging to Arthropoda  phyllum . The first is neuromodulation . The crustacean stomatogastrin network contains 30neurons while cardiac network contains 9neurons. They can act separately or together based on neuromodulators which are chemicals secreted either intrinsically or by extrinsic signals.Thus the neuropeptide RPCH allows conjoint action of both the networks by increased synaptic transmission between the two networks.

Similarly learning by conditioning and association can occur between nerve networks.The classical conditioning of a bell belonging to auditory circuit and salivation to gastric circuit occurs when ringing the bell initially in presence of food and later even in its absence causes salivation in dogs. The association of two circuits can only happen  if latent connections already exist between both the circuits as in dogs.

It is believed that most neural circuits in animals are innate. This requires a huge amount of information in the genes to configure them. Here the epigenetic mechanism comes to play where it can modulate the same circuit in various ways based on DNA methylation and acetylation by  non coding RNA in addition to alternative splicing of mRNA.  .The epigenetic mechanism gives predispositions to certain stimuli to the neurons in a circuit. The language module in humans has a predisposition to associate items to form word lists. It still requires working memory to form the association and learn language. The neural circuits characteristic of humans such as language .immitation and social capabilities all are innate in this manner.

Working memory and single episode learning

Both habituation and conditioned learning require repeated episodes of exposure for synaptic modification to take place. Higher animals with nerve nets capable of working memory can learn in a single episode. This occurs in both invertebrates as octopus ,insects and vertebrates. 

  Working memory activates part of a nerve network temporarily by reverberating circuits of neurons. While these neurons are in an activated state they can form sensory or actions associations more easily. Later on these associations are transferred to long term memory by cellular protien synthesis.The classic example of working memory in animals is navigation by hippocampal network

In the rat hippocampus a working memory for spatial map exists as a network of grid cells arranged in two dimensions. Each of the grid cell is activated to a smaller or larger extent depending on the position of the rat is confined space. The output of grid cells is fed into another network of cells the place cells. The place cells only respond to a particular coordinate position in the confined space.The position is the association of place cell with a landmark.So the place cells represent a relationship between landmarks. During movement the place cells are sequentially activated along the path taken by the rat.The memory of the path is transferred to LTM. Subsequently activation of this path by working memory retraces the original landmarks sequentially. This helps the rat in navigation as each activated place cell releases action towards next place cell and landmark.

   Navigation is an example of module specific working memory.Such memory exists in food storage  memory in honey bees and retrieval  of songs in birds.

Spectral time maps

The rewards and actions leading from them are organised as maps. The typical organization of a time map is in cerebellum. A conditioned stimulus (CS) from say PFC reaches the cerebellum by two pathways.The first path reaches cerebellum by parallel fibre to give excitation to nerve ends in purkinje cells. The purkenje cells output goes to deep cerebella’s nuclei and inhibits them.The parallel fibre have different strengths of stimulation to the purkinje cells.The other pathway ends in the deep cerebella’s nuclei.

    The unconditioned stimulus(US) activates the parkinge cells by climbing fibre s.This is called a teaching signal because if the CS signal and US signal coincide in time it weakens the parallel fibre signals to the purkinje cells while the rest of CS fibre s remain strongly excitatory.

When a CS occurs the purkinje cells are excited during its duration. However when the time of US inhibition comes the parkinge cells are inhibited.This in turn disinhibition the deep cerebella’s nuclei and it fires an action.

     The above mechanism enables a map to be created in cerebellum where different CS activates the action at different times.It also enables a top down control for different CS to be selected. The deep cerebella’s nuclei excitability can be matched to the required CS to be detected..A high excitable deep nuclei would get fired earlier. 

  Searching for rewards

Searching for rewards requires eye movement’s to orient towards a salient object. Normally a low level map in superior Colliculus  orients eye to salient features in the eye field. However in FEF we have a spectral time map for top down attention.Here CS has different timed attention based on the reward potential. A top down signal excites deep nuclei and changes it’s gain to match the reward signal most optimum at that moment

  Reward optimisation

Different rewards need selection based on internal needs of food,sex . Also rewards need to change in priority based on satiety. Rewards networks in OFC require a spectral time map. The deep nuclei for different rewards are in competition for their CS and action. Thus they mutually inhibit each other and the strongest wins. Moreover this strength of deep nucleus decreases with satiety.

Space maps 

    Similar to time maps there are space maps in the brain. The cerebral cortex has a laminar structure and this enables creation of space maps. Layer 4 receives sensory inputs like orientation of lines in simple cells . Since an object has a shape the corners require bipolar cells which have two oriented simple cells. The orientations undergo a loop from  layer 4 to layer 2/3 and then 6 And then layer 4 again. In this loop the various configuration of simple cells undergo competition and one survives. Layer 6 meanwhile receives a weak modular signal from below. It serves to select the competition by modulator signals to the strongest configuration.

    The modular signal from layer 6 can be utilised to use top down control from PFC to select an image by matching the strength of the winning configuration.

Sequence creation and similarity detection

In creation of a sequence the enhancement of a number of layer 6neurons have to be coordinated. A CS leading to an action must influence a subsequent CS to either enhance or inhibit it. This requires participation of reward networks which are spectral time networks along  with laminar cortical networks which code for spatial shapes.

     Similarity detection depends upon presence of neurons which detect change in orientation of lines or change in length of lines. The enhancement of matching orientation or length in layer 6 will detect similar lines in space maps. These maps at higher level can detect shapes and sizes of objects based on enhancement in layer 6 neurons.

 Reward maps which are spectrally timed are required to select sequences and similarity detection in spatial maps. Thus hunger leads to a search for fruits which have to be matched from memory. The search for food must follow a sequence so that the Same place is not repeatedly searched.

Vector configuration in brain networks

The spatial configuration in spatial networks form by selforganization,long range excitation and short range inhibition. Similarly the time networks form by Kalimantan filter networks. These cannot be analysed vectorial lay. However once self organization and learning has taken place the vectors of space or time can be triggered by probabilistic sequences. This resembles the rolling down of energy in a slope till the result is obtained. (Involuntary thinking)

Ack:Introduction to systems biology(exact adaptation in bacterial chemotaxis), Invertebrate learning and memory,Human epigenetics:How science works,Conscious mind resonant brain ,Working memory and language and modular mind,Mechanisms of memory