Microsystems for Bioelectronics

limited (see Table 6.5), and therefore harvesting of chemical energy from outside of the cell is

mandatory. For example, the E. coli cell contains up to 400 000 glucose molecules (Table 6.5) and each

glucose molecule can produce ~30 eV ~ 5  10

–18

J of energy (see Table 2.2.2 in Chapter 2). Thus,

based on Table 6.5, the total energy stored in cellular glucose in an E. coli cell is ~2  10

–12

J. Note that

in addition to glucose, energy is also stored in a number of derivatives of the glucose catabolism

process; therefore the total stored energy is somewhat larger, though it remains in the range estimated

above for glucose.

On the other hand, the power consumption of E. coli is about 1.4  10

–13

W(Table 6.5) and the cell

divides approximately every 2400 seconds (40 minutes – see footnote for Table 6.5). Thus the total

energy the cell uses during one division cycle is

E

cell

¼ 1:4  10

13

W  2400 s z 3  10

10

J (6.34)

When starving, E. coli consumes some of its ribosomes as an internal source of both energy and

nutrients (e.g. carbon and nitrogen) [43]. In fact, starvation of a population of E. coli cells results in

disintegration of some of the cells (lysis) within a few hours. Those cells that remain alive are then able

to utilize the remnant of dead bacteria to support their existence [44].

While at typical conditions occurring in nature E. coli does not have significant internal energy

storage, there are also exceptions. At some abnormal conditions E. coli can synthesize as much as 20%

of its dry weight as glycogen (glucose polymer) [54], which corresponds to ~10

–9

J of stored energy.

This amount of energy is equivalent to ~3x the cell cycle energy budget (6.34).

Many bacteria do not surv ive under long deprivation of nutrients. However, there can be

remar kable ex ceptions in whic h some bacteria have developed amazing mechan isms and optimi-

zation strategies to live under nutrient limitations [44–46]. Also cells can utilize other sources of

energy such as light as was briefly discussed in Chapter 1. Eukaryotic cells, on the other hand, which

are generally larger and more complex than bacte ria, can store a considerable amount of energy

internally.

6.10 BENCHMARK IN CARBO INFORMATION PROCESSOR

6.10.1 Top-down estimate of overall computational performance

As was discusse d above in Section 6.2.6, a conservative edge for estimated information content of

a bacterial cell such as E. coli is I

cell

~10

11

bits. In other words, for the correct placement of all

molecules within the cells, e.g. by the von Neumann’s universal constructor, 10

11

bits needs to be

generated by the cell processor. This quantity will be referred below as the binary information content

per task, I

task

¼ I

cell

, where ‘task’ refers to a completed assembly of a cell.

A measure of the information-processing rate can be made for the time required for a cell to

reproduce itself. A typical reproduction time of an E. coli cell is 2400 seconds (Table 6.5), and this is

taken to be time required for one computational task/cycle, t

task

. Thus the number of bits that must be

processed per second (bit rate), F

in carbo

, by the E. coli cell is:

F

incarbo

¼

I

task

t

task

w

10

11

2400

z 10

7

bit=s (6.35)

6.10 Benchmark in carbo information processor 177