Structure-Based Models of NO Diffusion in the Nervous System - Computational Neuroscience: A Comprehensive Approach - page 120

B C

100

0.25

75

0.2

50

0.675s

Threshold
0.625s
0.65s
0.675s

25

0.65s
0.625s

0.15

0

50 100 150 200 250 300 350 0 0.2 0.4 0.6 0.8 1

Distance ( µ m) Time (s)

Figure 4.7
Concentrations of NO at several time points during NO synthesis in a line through the centre
of ordered arrays of NO synthesising fibres of diameter 2µ m . B. NO concentrations generated
by 100 fibres separated by 36µ m after 0 . 625 , 0 . 65 and 0 . 675 s of NO synthesis. The dashed
line shows the threshold concentration. C. Area over threshold due to 100 fibres separated by
36µ m for NO synthesis of length 1 s plotted against time after synthesis. Here even though
there has been 600 ms of synthesis, just 100 ms more extends the affected region from virtually
nothing to over 50000µ m 2 .

there is a big advantage in using separated sources.
What about the temporal dynamics of the NO signal? Examining the time-course
of the NO signal generated by an array of 100 fibres spaced 36 m apart, we see a

delay until areas reach an above threshold concentration as we did for single sources
(Figure 4.7). This time, however, rather than the delay being for points outside the
source only, here there is a delay until any point is affected by NO, after which
there is a very steep rise in the volume affected. This is a common feature of sig-
nalling from dispersed sources because the summation of NO from several separated
fibres means that the concentration in and around them is, in a sense, averaged and
hence, smoothed. Thus due to the dynamics of diffusion one tends to get a relatively
even concentration within the synthesising region with small peaks around the fibres
themselves (Figure 4.7). In conjunction with the use of a threshold concentration,
this means that there will come a point when the concentration in a region around
the fibres is just sub-threshold and a small increase in the general level of NO will
result in large areas rising above threshold, as shown in Figure 4.7. We refer to this
feature as the interaction effect.
Again, the impact of the interaction effect will vary depending on the spacing used
and a large range of temporal dynamics can be seen ( Figure 4.8B) . In particular, the
delay before the start of interaction can vary from nothing to more than a second, with
the delay growing as the spacing is increased. This means that a system with optimal
spacing, in terms of extent of the affected region, will experience a considerable
delay before the region begins to be affected, with the total area affected suddenly
rising sharply at the end of the delay. This raises the intriguing functional possibility
of a system which is completely unaffected by NO for a given length of time (a

µ