Team:UT-Tokyo/Counter/Project/Project

From 2014.igem.org

Revision as of 12:46, 17 October 2014 by Yoichi (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

The word "counter" may remind you of the machine with which you can count the number of objects, such as persons and vehicles. Some people familiar with electronic circuits may remind of the logic circuit. In each case, the system is regarded as memory device that remember the number of inputs, which is important for our lives.

In the natural world, cellular counters also memorize the number of events. For example, there are telomere length regulation[1][2], cell aggregation[3][4], etc. Telomere length of Saccharomyces cerevisiae is regulated by the number of the Rap1 protein, indicating the existence of counting system. The cell aggregation size of Dictyostelium is regulated by counting factor (CF). CF counts the number of aggregating cells and negatively regulates the cell adhesion. In these ways, cellular counters are widely utilized for the regulation of biological systems.

With synthetic biological approach, Ari et al constructed a cellular counter termed the riboregulated transcriptional cascade (RTC) counter[5]. The state transition occurs after an arabinose induction (Fig. 1). The system is regulated by riboregulators. The Biobrick part of this cellular counter has already existed, which was constructed by Tokyo-Nokogen 2009 and was named BBa_K225002, BBa_K225003 [6].

In order to expand the function of this counter, we added "reset system". The reset system enables the transition from any state to the initial state after a particular input. The property is expected to apply for a deterministic finite automaton, which is the system developed from information science. Within the finite number of states, the system makes the transition to another state in response to a particular input.

As the key of its resetting mechanism, our counter utilizes the regulation system based on sigma factor and anti-sigma factor. Sigma factor is a subunit of RNA polymerase and help it bind to the specfic sequence of the promoter. Anti-sigma factor blocks sigma factors from interacting with RNA polymerase. The utilization of this regulation system brings about benefit for attempt to make automaton.Firstly, the number of states can be increased easily because what we have to concern is the combination of sigma and anti-sigma factor. Secondly, the crosstalk between sigma and anti-sigma factors can be circumvented even if you raise the number of states. That is because we can choose such combinations between sigma and anti-sigma factors that have little crosstalk.

Fig. 1 The concept of RTC counter. After first, second and third induction of arabinose, the state of cells moves from 0 to 1, 1 to 2 and 2 to 3, respectively.

Sigma factor is a subunit of RNA polymerase related to promoter recognitions. There are two types of sigma factors, one that is housekeeping (expressed constantly) and another that is expressed under some conditions. It is possible to deliver genes of sigma factors that are not derived from some species to another species, so we can use many kinds of sigma factor. Which promoter RNA polymerase tends to bind to is decided by which sigma factor RNA polymerase is bound. Especially some sigma factors promote only transcription of a specific promoters. Therefore if we choose a set of sigma factors and promoters skillfully, we can controll transcription without crosstalk. Here, “without crosstalk” means every sigma factor in the set promotes only transcription of the cognate promoter.

<img src = " Nakashima_image%281-0%29.png " class = "figure-height" /> <legend>Fig.1 sigma factor</legend>

Anti-sigma is also a protein which is related to transcriptional control by sigma factors. Anti-sigma prevents sigma factors from binding to RNA polymerase. Consequently, sigma factor cannot activates transcription of cognate promoters. Similarly as sigma factor, there are many kinds of anti-sigma and some anti sigmas prevent only specific sigma factors from transcriptional control. Therefore if we choose a set of sigma factors, anti sigmas, and promoters skillfully, we can activate control, not only activate but also repress, transcription without crosstalk.

Transferred sigma factors or anti sigmas may have negative effects on the growth of the host cell. For example, anti sigma may prevent housekeeping sigma factors from transcriptional control. However it is no problem since not all sigma factors and anti sigmas have negative effect and there are many kind of sigma factors and anti sigmas.［1］

sigma-memory construction

This is the construction of our sigma-memory.This gene circuits is composed of three parts. The gene of a sigma factor is placed at the downstream of the promoter that is induced by a substance A and at the downstream of Psigma, which is induced by the sigma factor. The gene of the cognate anti sigma is placed at the downstream of promoter which is induced by a substance B.

At first, there is no sigma factors. If input A exists, sigma factor is expressed. Then the positive feedback circuits of sigma factor starts producing sigma factor, and consequently sigma factor will remain. Though if input B exists, anti-sigma is expressed and the positive feedback circuits is inhibited. Both sigma factor and anti-sigma are subjects to degradation［2］, so all of them are decomposed after some time and sigma-memory returns to its original state.

We can regard existence/absence of sigma factor as 1/0 of memory, and this values of memory can switch by input A or input B. Using the promoter which is cognate to the sigma factor, the information whether the value of memory is 1 or 0 can be derived. For example, consider the circuits on the right. The reportor is expressed if and only if sigma-memory's value is 1 (i.e. sigma factor exists).

In addition, no crosstalk sets of sigma factors, promoters, and anti-sigmas enable us to make multi-sigma-memory gene circuits. To make the explanation easier, consider the case in which E. coli has two sigma memories, sigmaA-memory and sigmaB-memory. SigmaA The value of sigmaA-memory change from 0 to 1 if input A1 exists and change from 1 to 0 if inputs B1 exists. Also the value of sigmaB-memory change from 0 to 1 if input A2 exists and change from 1 to 0 if input B2 exists. If the four input A1, B1, A2, and B2 have no crosstalk, sigmaA-memory and sigmaB-memory also have no crosstalk. For example, when input A1 exists, only the value of sigmaA-memory change from 0 to 1 since sigmaA promotes only transcription from PsigmaA (promoter that is cognate to sigmaA). Since the transcription from PsigmaB is not activated, sigmaB is not expressed and the value of sigmaB dose not change. The same is true of input A2. When input B1 exist, anti sigmaA is expresses and the value of sigmaA-memory changes from 1 to 0. However, anti sigmaA has no effect on the transcription of PsigmaB and the value of sigmaB dose not change. The same is also true for the input B2. So it can be confidently said that E. coli has two sigma memories.

Since this construction has a positive feedback, leakage of promoter may be a serious problem. If the promoter has leakage and sigma factor is expressed when input A dose not exist, this error may be enlarged by positive feedback. (Whether the error is really enlarged depends on whether the leakage of sigma factor is large compared to the degradation of sigma factor.) However if the leakage of anti-sigma is considerably large, sigma factor dose not produced from positive feedback when leak of sigma factor occurs. Therefore leakage is not an obstacle of our project if we choose the promoter skillfully.

sigma-memory and automaton

Speaking easily, an automaton is referred to what has states and transition rules. Every state has a transition rule, say it is decided what is the next state if an input exists. For example, the automaton on the right has five states, 0, 1, 2, 3, 4, and transition rules are represented by arrows. If the current state is 0 and the input is b, the next state is 2. Transition rules having loops or feedbacks are allowed.

sigma-memory can be considered an automaton. Every tuple of the value if sigma-memory (say, (1, 1, 0, 1, 0)) corresponds to a state, and the promoter which is used in the construction of sigma-memory decides transition rules. It is possible to use the value of one sigma-memory as an inputs of another sigma-memory. This fact enables us to make complicated transition rules.

For example, consider the case where E. coli has two sigma-memory. The value of sigmaA-memory( a sigma-memory which uses sigmaA in the construction as a sigma factor) change from 0 to 1 if a substance A1 exists, also the value of sigmaB-memory change from 0 to 1 if the value of sigmaA and a substance A2 exists. (This condition can be written in terms of mathematical logic, namely AND(sigmaA, IPTG)) Gene circuits which work as logic circuits (says AND, OR, NAND, etc...) have been designed.［3］ Therefore such gene circuit can be made. The value of sigmaA-memory changes from 1 to 0 when a substance B1 exists and that of sigmaB-memory changes from 1 to 0 when a substance B2 exist. In this case, it is characteristic that sigmaB-memory is affected by sigmaA memory. These gene circuits can be considered as the automaton on the below.

resettable counter by sigma-memory

Resettable counter is a device that can count the number of induction events of arabinose, and expresses a reporter correspond to each states. In addition, the count can be reset by IPTG induction. Resettable counter can be considered as an automaton. The automaton has two inputs, input A and input B. Each states corresponds to the count, namely the states of the automaton is 0,1,2,...etc. The transition rule is very simple. When the current states is n and the input is A, the next states is n+1, and when the current state is n and the input is B, the next states is 0. Since a counter is one kind of automata, sigma-memory can be applied for making resettable counters.

2-counter is a counter that can count up to 2, and is an automaton that has three states, 0, 1, and 2. This automaton can be constructed by sigma-memory as following. Two kinds of sigma factor are necessary for constructing the 2-counter, so for the convenience we will call these two kinds of sigma factors sigmaA and sigmaB. In state 0, both sigmaA and sigmaB do not exist (i.e. sigmaA-memory and sigmaB-memory is 0). In state 1, only sigmaA exists and in states 2, both sigma factors exist. The input is the induction of arabinose or IPTG, and the transition rules with it are showed in the Figure.

A gene circuit that realizes this automaton can be represented as following. The value of sigmaA changes from 0 to 1 after arabinose induction (i.e. sigmaA is expressed when arabinose exists). This change corresponds to the transition from state from 0 to 1.The value of sigmaB-memory changes from 0 to 1 when both arabinose and sigmaA exist. The conditional branch can be made by using AND function. This change corresponds to the transition from state 1 to state 2. Both the value of sigmaA-memory and sigmaB-memory change from 1 to 0 after IPTG induction (i.e. anti-sigmaA and sigmaB is expressed when IPTG exist).

riboregulator

According to the discussion in 2-1, it is necessary to use AND function for the construction of a resettable counter. Therefore a riboregulator is used for the realization of the AND function.

A riboregulator is a post-transcriptional regulation system composed of two kinds of RNAs, Cis-repressed mRNA (crRNA) and trans-activating RNA (taRNA). CrRNA forms a stem-loop and its ribosomal binding site (RBS) is covered. Consequently the gene coded in cis-repressed mRNA isn't translated. However if trans-activating RNA exist, crRNA and taRNA are hybridized and RBS gets exposed and translation starts.［4］

Namely, the gene coded in crRNA is expressed only when both crRNA and taRNA is transcripted. Therefore this system can be considered as the AND function. For example, consider the gene circuit, Plac-taRNA-d.term-Pbad-cr-RBS-GFP-d.term.(cr is the sequence in crRNA which binds to RBS) If and only if both IPTG and arabinose exists, GFP is expressed. Therefore, the riboregulator can combine two promoters and produce AND functions

resettable counter construction

The construction of our resettable counter is explained here.

Ecf20_992 (Sigma20) and Ecf11_3726 (Sigma11) are used as sigmaA and sigmaB as mentioned above respectively. These two sigma factors strongly activate the cognate promoters, and their inhibition of growth is negligible. ［1］The promoters correspond to sigma20 and sigma11 are Pecf20_992(Psigma20) and Pecf11_3726(Psigma11), respectively. The sequences of these two promoters has no restriction site of Ecor1, Xba1, Spe1, and Pst1. The cognate anti-sigmas are anti-20 and anti-11 respectively. These two anti-sigmas considerably prevent the cognate sigma factors from activating transcription, and their inhibition of growth is also negligible. These sigma factors, cognate promoters, and cognate anti-sigmas has no cross talk.

Cis-repress sequence is crR12 and trans-activating RNA is taR12. This pair is selected because the leakage is small. ［4］The reporter in this construction is GFP.

mechanism and extension

At first, both the value of sigma20-memory and sigma11-memory are 0. Only the crRNA coding sigma20, which is at downstream of constitutive promoter is translated. After the first induction of arabinose, taRNA at the downstream of PBAD is transcribed and the crRNA coding sigma20 is translated, and the value of sigma20-memory changes from 0 to 1. Since sigma20 exists, crRNA coding sigma11 at the downstream of Psigma20 is transcribed. After the second induction of arabinose, taRNA is transcribed and sigma11 is expressed, and the value of sigma11-memory changes from 0 to 1. At this time, GFP at the downstream of Psigma11 is expressed and we can check the count as 2.

After an induction of IPTG, anti-sigma20 and sigma11 are expressed, and the value of sigma20-memory and sigma11-memory will be 0. Therefore the count is reset.

Since there are many kinds of pair of sigma factors and cognate promoters which have no crosstalk, n-counter can be made by the same way. Besides, the construction of 2-counter can be simplified.

This simplified counter is made by only one sigma factor. It works as the same way as original 2-counter unill 1 count. After 1 count, crRNA coding GFP is transcribed at the downstream of Psigma. Therefore when the next arabinose induction occurs, GFP is translated. This expression can be considered a report of 2 count. This simplified counter can be also extended to n count. Simplified counter can count up to larger numbers compared to the original counter even when the same number of sigma factors are used, but cannot be reset from their final count. We did experiments on this simplified counter.

comparison with previous counter

Our sigma-Recounter is improved version of the previous counter constructed by Ari.［5］In the previous counter, T7 RNA polymerase and T3 RNA polymerase is used, while in our counter sigma factors is used. Using sigma factor has two merits. One is the ease of extension for n-counter. The number of RNA polymerase derived from virus is limited, but sigma factor has great divergence. Consequently it is more easy to construct n-counter by using sigma factor. Another is the existence of inhibitor. Anti-sigma is inhibitor of sigma factor which has no cross talk. Inhibitor is necessary to realize reset function.

Other difference is the positive feedback circuits. Previous counter has no feedback circuits. Since sigma factor is more subject to degradation than RNA polymerase, positive feedback circuits is necessary to keep "memory" (i.e. for sigma factor to remain) in our counter.

As described above, the genetic circuit we constructed can be considered as an automaton. In a previous study[1], many sigma and anti-sigma that regulate transcription without crosstalk have been reported. Thus, an automaton that has many states can be constructed. Furthermore, though in this project reset is transition from other states to state 0, more general system that is capable of changing one state to any other states is possible. Thus more general automaton by genetic circuits may be possible. Examples of an automaton are such as biocomputer or lifegame etc. Bringing this concept from algorithmic world into synthetic biology is our challenge. Though in our project input is a single short intermittent signal, a more general circuit that responds to more general input is possible to be considered by integrating additional circuits into our circuit.

For example, the change of balance between 2 substances itself can be considered as a input. Considering substances A and B, this additional circuit is possible:

pA-repressorB-reporterA-activatorX-pX-repressorA

pB-repressorA-reporterB-activatorY-pY-repressorB

Here, substances A/B activate promoter A/B. This additional circuit essentially contains toggle switch structure with delay negative feedback loop. For example, when substance A become dominant against substance B, the toggle switch amplifies the dominance of promoter A and the following negative feedback suppress the dominance. Consequently, this additional circuit is expected to convert the change of the dominance to a pulse expression of reporter protein. Therefore, this additional circuit can expand the range of input. As this circuit can be a monitor of a milieu if A is industrial waste, our genetic circuit can be applied to much wider range of problems.