BIOSENSOR PROJECT

Roi Leibowitch

Shahar Kvatinsky

Idith Farkas

Instructors: Prof. Aharon J. Agranat, Harel Ilan

Picture of the system

2

Contents

Abstract 3

Chapter 1 - Previous work 4

Chapter 2 - Biological Background 6

Chapter 3 - The system 10

Chapter 4 - Experiments Summary 16

Chapter 5 – future work 28

Summary 31

References 32

Appendix 1 – Data sheets 33

Appendix 2 – Labview control program 37

Appendix 3 – NI CompactRio Spec 40

Appendix 4 – Minimal medium MOPS glucose 43

Appendix 5 –An experiment protocol 44

3

Abstract Early warning of chemical hazards that occur in water reservoirs and supply systems is

becoming a key issue in maintaining water quality and safety in urban and rural

environments.

We present an integrated system for realtime inline monitoring of pollutants in water,

based on E-Coli bacteria.

The core of the system is genetically engineered E-coli microbes that respond to the

presence of the pollutants by producing a dye material. The amount of the dye produced

is in turn measured in an opto-electronic detection device which consists of a LED and a

detector. The system includes an automatic sampling apparatus consisting of 10 cuvettes1

in which the bacteria, dried, lie until a sample is taken into a cuvette from the reservoir.

Up to two hours from the time a request is made to sample the water, a result stating

whether the water is toxic or not is available then transmitted via WiFi to a base station.

1 A type of square test tube designed to let light pass through it for optical testing

4

Chapter 1 – Previous work In this section we will present the previous work done in the field of water toxicity

analysis (WTA).

It is common to divide WTA to two separate fields, chemical and biological. The

difference is in the discovery mechanism. In the chemical mechanism, the response to a

certain chemical is measured, while when dealing with biological mechanism, a reaction

of a biological element is measured.

As for today, both of the options above usually require manual sampling of the water

source, and transferring the sample to a central lab, where the process of examining the

water might take up to 24 hours. Obviously, this time is too long when we deal with

water pollution. In addition, the cost is several hundreds of dollars for each sample.

The chemical analysis of the tested sample is a more mature technology, and is used to

identify specific toxic chemicals or biological elements, from a closed list of materials (as

opposed to biological detection, that can identify a general problem, Meaning to react to

unknown toxins in a sample, as can be done with bacteria, more on this subject will be

explained in the next chapters).

The biological mechanisms can be divided into a number of categories, based on the

reaction of the bacteria2:

• Luminescence - the bacteria are emitting light.

• Fluorescence – the bacteria are creating fluorescence material.

• Colorimetric – the bacteria cause a change in the color of the environment.

A great number of academic and industrial researches were performed in this field [not

only in regards to water toxicity but for general detection as well] here are some of the

main actions:

• TAU water toxicity research group - “Tel-Aviv University – Toxicity

Measurement System” (TAU-TMS) was a solution for working with fresh whole

cell biosensors, however wasn’t sensitive enough to work with rehydrated

biosensors. This system used fluorescence bacteria. Another TAU system was

based on luminescence bacteria [ 8, 2].

2 Elaboration on this terms is in the Biological Background section

5

• Toshiba Corporation has developed essential technology for an advanced

biosensor chip capable of wide-ranging applications. Such system can measure

glucose in blood with up to 500 times the sensitivity of current biosensors. That

new biosensor comprises of a sensor chip that interacts with the sample and a

detection section that records the result [ 2].

• PointSource Inc. has developed a cost-efficient testing apparatus that will allow

for continuous, real time monitoring of the water supply for the presence of

dangerous pathogens. The patented system relies on a laser beam to identify

microorganisms in water. A side stream of water, which assumed to be

representative of the water in general, passes through a laser beam. As it passes

through the beam, light is scattered. The system gathers the light and, using a

mathematical technique, can determine what kind of particle it is by its shape, size

and internal composition [ 2].

• Cyanide detection: Colorimeters test kit includes reagents that when added to a

water sample, react with the available cyanide ion to form a colored solution. A

vial containing this solution is inserted into the hand-held colorimeter that

measures the intensity of the sample's color and reports the cyanide concentration

[ 2].

• South Korea Advanced Environmental Monitoring Research Center - multi-

channel continuous water toxicity monitoring system was implemented to sample

water discharged from power plants in order to detect and classify their toxicity,

using several recombinant bioluminescent bacteria. Each channel of the system is

composed of a series of two mini-bioreactors to enable a continuous operation,

i.e. without system interruption due to highly toxic samples [ 5].

As for now, there is no system which is based on colorimetric biological detection,

although this field is well known for biological measurement. There is a lot of biological

data on colorimetric detection, but not from engineering perspective.

Another issue that can be noticed is that there is no biological water sampling system

which is continuous and stand alone.

6

Chapter 2 - Biological Background The basic concept is to use bacterial reporter strains that are genetically engineered to

react to different types of selected toxins. The reaction can have optical characteristics

and is dependent on the bacterial strain and the level of the toxin concentration.

The concept behind the engineering of the reporter strains is the fusion of the DNA

sequence of a gene promoter, known to be activated by the presence of genotoxic chemicals,

to a gene or a group of genes the presence or activity of which can be monitored

quantitatively, preferably in real time. After that, whenever the bacteria are exposed to this

material, the promoter will initiate transcription and protein formation.

We compared several types of bacteria's reactions; each reaction is involved with

different type of protein:

1) Fluorescence bacteria – The bacteria’s reaction creates a fluorescence protein, that

is excited (absorbed) at a unique wave length and emit at longer wave length.

In the world of Biosensors the drawback of the fluorescent proteins is the fact that

there is no amplification of the signal- one protein=one excitation, not like with

enzymatic reactions in which one protein can create 100 products that produce a

signal.

Common proteins - Red Fluorescence Protein (DsRed), green fluorescence protein

(GFP)

Advantages - very stable, lasts for days.

Disadvantages – initial reaction time is slow (reaction starts after 60-100

minutes), not sensitive and has long time intervals for a substantial reaction.

2) Luminescence bacteria – The bacteria’s reaction is creating a luminescent protein,

the bacteria are producing a proteins complex that enables the enzymatic reaction

in which photons are released. The creation of the reaction consists of two

processes – 1. The synthesis of the enzymatic complex that produces the substrate

and creates the luminescent reaction. 2. The enzymatic reactions in which the

substrate is produced and photon is released.

Common proteins - luxCDABE

Advantages - fast detection (after 30-60 minutes), sensitive signal detection.

Disadvantages – fade after ~3 hours, low intensity.

7

3) Colorimetric bacteria – The bacteria’s reaction is creating an enzyme, that when it

reacts with colorimetric substrate it creates a colorimetric product. The product

absorbs light at a specific wave length.

Common colorimetric product - p-nitrophenol (PNP), ortho-nitrophenol (ONP).

Advantages – fast initial reaction and fast detection (after ~30 minutes),

continuous enzymatic activity (the process is continuous and therefore keeps

producing the protein).

Disadvantages – no relevant data for engineering purposes (never been used for

such purposes), there is a spontaneous reaction even without a toxin (slower

reaction), toxins with color may influence the readings.

Figure 1 – Colorimetric detection

In figure 1 there is a schematic of the detection process, as explained above.

We decided to use colorimetric bacteria, due to its advantages.

Substrate

There are several types of colorimetric substrates:

1) para-nitrophnyl phosphate (PNPP) – The reaction of the substrate with alkaline

phosphatase (type of an enzyme) produces the product p-nitrophenol (PNP), a

yellow material that its absorption spectrum is as seen at figure 2 (center of

absorption is 405nm). More soluble than ONPG.

8

Figure 2 -– PNP absorption (different PH) [1]

2) Ortho-Nitrophenyl-β-galactoside (ONPG) - The reaction of the substrate with β-

galactosidase (type of an enzyme) produces the chromogenic molecule ortho-

nitrophenol (PNPG), center of absorption 420 nm.

We characterized the substrates (see experiments chapter) and decide to use PNPP.

Drying of the bacteria:

Different types of storage of the bacteria:

1) Liquid – a solution of substrate and bacteria. The fastest way for reaction (the

bacteria is alive). In order to be preserved for longer time (several weeks) and to

prevent bacteria reaction (until sampling) it needs to be cooled (4°C).

2) Solid firmly fixed – the bacteria is fixed on an agar (complex sugar), which lets

the liquid to penetrate into the bacteria, and also requires cooling.

3) Dried – drying the bacteria, the substrate and the medium into powder. Can be

stored for a long time (1-2 years). Needs vacuum or a very dry environment. It is

better to store the powder at low temperature. Takes more time to “rehydrate” the

bacteria (another ~20 minutes). 50% of the initial live bacterial cells die during

the process.

9

We decided to use dried bacteria in order to get an independent machine that can be

placed near a source of water for a long time, without the need of maintenance.

The drying process:

We mixed together the correct concentration of bacteria and PNPP, including MOPS (the

medium, water with sugar, salts and amino acid, see appendix 4) and threalose (a sugar

that helps the bacteria to survive the drying process).

The drying process is actually lyophilization process.

The drying stages:

1) Freezing – in order to stop the biological processes cooling the solvent to -40°C.

There is a faster process which is to dip the solvent in liquid Nitrogen. We didn’t

use this process.

2) Vacuum – in order to take out the water vapor.

3) Temperature rising – as the water vapor is taken out, the pressure becomes lower.

Therefore we raise the temperature in order to keep the process of taking out the

water vapor – the process is finished at 30°C and lasts about 2 days.

Since it was the first time that Professor Belkin’s lab dried bacteria in cuvettes, we had to

deal with several aspects:

1) Cuvettes temperature proof at low temperature – we cooled the plastic cuvette to

-80°C and saw that it can withstand such low temperature.

2) Find a way to seal the cuvette - we ordered special taps, which allow us to keep

the dried powder under vacuum.

3) Mixing substrate and bacteria together in the drying process.

4) The cuvettes are made from plastic and therefore have worse heat conductivity;

we cope with it by using Copper small plates that we coupled to the cuvettes.

10

Chapter 3 - The system In this section we will present the system, its various parts and their way of work.

The system we built is capable of detecting the presence of genotoxicants in water

sampled from a reservoir and transmitting its findings to an off location base station. We

used the substance Naladixic acid (NA)3 as a model toxin. In our system, as stated in the

Biological Background section of this document, genetically engineered E. coli reporter

strain respond to the presence of the NA by altering the color of a substrate material.

NA was used as a model toxin to characterize our system and calibrate it, but any of a

large number of substances can be detected using our system with the need only to

recalibrate it. In such a case the use of a different strain of bacteria, engineered to respond

to another toxin should be used.4

The biological reaction takes place inside cuvettes5. The bacteria are dried in a process,

also described in the Biological Background section of this document. The dried bacteria

stay in the cuvettes under vacuum untouched for up to months until the water is sampled.

The cuvettes are part of a motorized stack, in the case of our prototype containing 10

cuvettes. When all the samples in the stack are finished, the stack needs to be replenished.

The conclusion stating whether there is NA present or not is made according to a

predetermined toxin level under which we state the water isn’t toxic and above it that it

is. It takes up to two hours to get a clear reaction from the bacteria and reach the

conclusion. The reference level of toxin was determined based on the series of

experiments we conducted on both dried and non dried bacteria using different levels of

NA. These experiments are presented in the Experiments Summary section of this

document.

3 Type of antibiotics 4 See Future Work section for further details on the possibilities of using such bacteria 5 Type of testing tube used to put light through the sample inside it

11

Figure 3: Scheme of the system6

The biological reaction

Originally the substrate used in the biological reaction is PNPP – a colorless material.

Once the bacteria are mixed in with the NA and the PNPP, the latter turns into PNP

which is yellow. As the concentration of NA is higher, so is the concentration of yellow

6 Whenever a connection is marked by D17 for example, it is connected to port 17 of the DAQ

12

material in the mixture. It is the amount of color produced that our system, in fact,

detects.

Still, we take into account that this reaction occurs also when there is no NA present,

since the enzyme responsible for it is produced by the bacteria at all times, in basal levels.

For this reason all reaction is measured against the reaction seen with no NA at all,

referenced in this document as a concentration of 0 PPM NA.

The opto-electronic system

A LED of 400 nm7 was picked so that its spectrum matches that absorbed by the yellow

product – PNP. Figure 4 and figure 5 show that match.

Figure 4: Emission spectrum of our LED

The light of the LED is concentrated using an objective lens and then illuminates a

cuvette containing bacteria, PNPP substrate turning into PNP and a sample of water from

the reservoir. At the other side of the cuvette, a visible light photodiode detector8 collects

the light that was not absorbed by the sample and changes its reading accordingly.

The detector's reading in volts is received by a DAQ card and processed in a computer.

7 See appendix1 for the data sheet of the LED 8 See appendix 1 for the detector’s photodiode’s datasheet

Figure 5: PNP absorption (different PH)[1]

13

The sampling system

10 cuvettes are aligned in a motorized stack. Each cuvette is held in place by a copper

mount and separated by an insulating material from the others. The copper mount eases

heat flow from a Thermo-electric cooler (TEC) placed under the cuvette used for the

current sample. The TEC is used to heat the sample to 37ºC – an ideal temperature for the

bacteria's reaction.

A new cuvette comes to its place in the opto-electronic system whenever a new sample is

requested. The TEC moving under them is placed on a spring so that it is pressed against

the copper mount for maximum contact and heat flow. The mounts are curved at the

edges to facilitate the TEC's movement across them.

A pump with adjustable speed pours water into the current cuvette. At the end of the pipe

there is a syringe and a needle. A stepping motor connected to the syringe moves it down

to puncture the stopper of the cuvette, then, the water flows through the syringe into the

cuvette.

Before each sample is taken, 0.5 liter of water from the reservoir is pumped through into

a sink, which is part of the cuvette stack, to remove any leftovers of the latest sample in

the pumping system.

To stir the sample, mini magnetic stirring rods placed inside each cuvette and a magnet

outside the cuvette are used.

Figure 6: A picture of the opto-electronic system

14

Controlling and processing

All instruments of the system are controlled via computer using a DAQ card and a

Labview program shown in appendix 2.

The result, stating if there is toxin present or not, is transmitted via e-mail to an off

location base station.

When a request is made from the base station to take a sample the following commands

are given to the system:

Figure 7 : A sketch of the cuvette stack used in its planning

Figure 8: A picture of the cuvette stack

Figure 9: A cross section of the cuvette stack showing the TEC under the current cuvette

mounted on a spring for maximum contact

15

0. Move to sink

1. Puncture the cuvette stopper

2. Set pump speed to 4.35 V (speed 200 RPM)

3. Activate pump

4. Wait 4 minutes to clean the pipes

5. Deactivate pump

6. Get out of stopper

7. Move to sample location, set TEC to 37°

8. Puncture the stopper

9. Set pump speed to 1.407 V (speed 10 RPM)

10. Activate pump

11. Wait 4 seconds to pump 0.6 ml of water

12. Deactivate pump

13. Get out of stopper

14. Stir

15. Activate LED

16. Wait 2 hours and collect data

17. Compare to reference toxin level and send message

18. Deactivate LED and turn off TEC

16

Chapter 4 - Experiments Summary All the experiments were made on bacteria that were genetically engineered to react to

Nalidixic acid (NA) as the toxin. The genetically engineering was done in Prof. Belkin

Lab.

The experiments’ steps:

1. Optimization test – substrate – 12/3/09.

2. Optimization test – toxin concentration - 25/3/09.

3. Optimization test – bacteria concentration – 24/6/09.

4. Kinetic test (reaction over time) – 1/4/09, 13/5/09, 20/5/09, 1/6/09.

5. Dried Bacteria characterization – 29-31/7/09, 25-27/8/09, 1/9/09-3/9/09,

6. Heating characterization – TEC vs. incubator – 31/7/09, 4/8/09.

7. Full system characterization – 8/9/09, 10/9/09, 17/9/09.

The optimization tests were done on each component in the biological part (substrate,

bacteria and toxin). We made those tests using liquid bacteria and characterized the

optimal substrate (PNPP vs. ONPG) concentration, the optimal bacteria concentration

and the reaction of the bacteria to different toxin concentration. From those experiments

we determined the substrate and bacteria concentration and the reference voltage for

the bio sensor system.

In the kinetic assay we monitored the reaction over time, this way we could determine the

sampling time of our system.

The assay using the dried bacteria was made in order to find the differences between

dried and liquid bacteria and to get a more accurate determination of the bacterial

concentration, the reference voltage and the sampling time.

In the heating characterization we determined the correct TEC temperature in order to

get 37°C inside the cuvette. We also compared the reaction inside an incubator and in a

cuvette coupled to a TEC.

In the full system characterization we performed a full running of our system. We

determined from those experiments the time and speed of the pump, considered the

steering issues and checked the full integration of the system.

17

Note – different experiments were made in different dates and different conditions (new

bacteria, different equipment etc.), therefore we can't compare the absolute results of one

experiment to another and can only compare the trend. The parameters for the final

system are determined in the last experiments on the system in its final setup.

The experiments:

1. Optimization test – substrate:

Date: 12/3/09.

Goal: decide which type of substrate to use (PNPP vs. ONPG).

Description: we compared the colorimetric product of PNPP (PNP) vs. the

colorimetric product of ONPG (PNPG). We wanted to see the concentration of

the product in the water and the values that were measured in this concentration

(the measured voltage, which implies on the absorption of light).

We determined the height of the water in the cuvette (different volumes in similar

cuvettes) and how the measured voltage depends on it, in order to determine the

optimal water volume for sampling (minimum volume that can collect all the light

from the LED).

We validated that presence of PNPP and ONPG keeps the water clear (no change

in the measured voltage).

Results: we saw that PNPP is much more soluble than ONPG and therefore made

more tests on it.

The results of the measured voltage in different concentration of PNP are shown

in graph 1.

18

PNP concetration test

0

1

2

3

4

5

6

7

0.01 0.1 1 10 100

PNP concetration [mg/ml]

Vo

ltag

e [V

]

Graph 1 – PNP concentration test

Conclusions:

a. Both substrates (PNPP and ONPG) keep the water clear.

b. PNPP is much more soluble and therefore is preferred.

c. Even a small amount of PNP is noticeable; there is a significant difference

between 0 and 0.05 mg/ml.

d. The optimal volume of water in the used cuvettes is 0.6 ml.

2. Optimization test – toxin concentration:

Date: 25/3/09.

Goal: characterization of the bacteria's reaction to different toxin concentration.

Description: we mixed the bacteria with the medium, the substrate and various

concentrations of the toxin (NA) and incubated it for one hour and a half. Then

we measured the reading in our detector 20 and 40 minutes after the incubation

(left in room temperature).

We put in every cuvette 0.3ml MOPS G (the medium), 0.3ml distilled water and

0.08% PNPP.

The bacteria OD (optical density) were 0.25.

We added 2 sets of cuvettes with different NA concentrations (0.15, 0.3, 0.6, 1.25,

2.5, 5 and 10 ppm) and also 2 control cuvettes (0 ppm of NA) and a blank cuvette

(a cuvette with NA but without bacteria).

The biological experiment protocol is in Appendix 5

Results: the results relative to the control (no NA) and in absolute value in volts

minus reading of control after 20 minutes are presented in graph 2 and graph 3.

19

20 minutes afer incubation, relative to 0 PPM

0

0.5

1

1.5

2

2.5

3

0 2 4 6 8 10 12

NA concentration [PPM]

Rea

din

g r

elat

ive

to 0

PP

M

Maximum value

Minimum value

20 minutes afer incubation, subtracted from 0 PPM

0

1

2

3

4

5

6

7

8

0 2 4 6 8 10 12

NA concentration [PPM]

Rea

din

g s

ub

trac

ted

fro

m 0

PP

M [

V]

Maximum value

Minimum value

We tested two samples of each NA concentration, which produce 4 data points of

results relative to 0PPM NA since there are two 0PPM samples as well. In the

graphs, maximum and minimum values refer to the maximum and minimum of

those 4 data points.

Graph 2 – Different NA concentration relative to 0 ppm

Graph 3 – Different NA concentration subtracted from 0 ppm

20

Voltage reading vs. bacteria concentration with close detector

0

5

10

15

20

25

0 0.2 0.4 0.6 0.8 1 1.2

bacteria concentration [OD]

det

ecto

r re

adin

g [

volt

]

10 ppm, A

2.5 ppm, A

0 ppm, A

10 ppm, B

2.5 ppm, B

0 ppm, B

Conclusions:

a. There is a significant difference between no pollution (0 ppm) and

presence of toxin (at least 1.5 V difference was measured).

b. The optimal reaction was measured at 2.5 ppm; there is a decline in

the reaction in a more concentrated toxin (due to bacteria's death

from the toxin).

3. Optimization test – bacteria concentration:

Date: 24/6/09.

Goal: determine the optimal bacteria concentration.

Description: we measured the response after an hour and a half in an incubator

with 4 different concentrations of bacteria (OD of 0.1, 0.25, 0.5 and 1), each with

3 different concentrations of NA (0, 2.5 and 10 ppm).

We also made two separate measurements with our detector placed at different

distances from the cuvette, each more suitable for a different range of voltages

corresponding to different concentration of bacteria.

Results: the results for the closer detector present in graph 4.

Graph 4 – The measured voltage for different bacteria concentration

21

Conclusions:

a. For an OD of 0.25 the results are best, with greatest separation

between various NA concentrations.

b. The detector placed closer, fits bacteria concentration of 0.25 OD

better, giving a voltage range of 15 volts between 0 ppm and 10 ppm

of NA.

c. For greater OD we get great dispersion of the light giving less

reliable results, since they can vary with changes in the cuvette

angle, and anyway the difference of voltages is not better than in the

0.25 OD case.

4. Kinetic test (reaction over time):

Date: 1/4/09, 13/5/09, 20/5/09, 10/6/09.

Goal: testing the reaction upon time and approximate the time for optimal

measurement of the level of toxicity.

Description: we prepared the cuvettes as mentioned in Appendix 5 (2 samples for

each NA concentration – 0, 0.3, 0.6, 1.25, 2.5, 5, 10 ppm), we put all cuvettes in

an incubator and tested it every 20 minutes.

We used OD 0.14 on 13/5, OD 0.22 on 20/5 and OD 0.238 on 10/6.

Results: we had difficulties to get good results for those experiments. The

experiments on 1/4/09 and 13/5/09 were a failure. There are several options for

the reasons of the failures, but we couldn't define the exact reason:

• Contamination of the samples

• Temperature changes

• Low OD

The result on 10/6 for all concentrations is presented in graph 5.

The result on 10/6 after 2.5 hours is presented in graph 6.

22

Kinetics Experiment Results

0

2

4

6

8

10

12

14

16

18

0:43:12 0:57:36 1:12:00 1:26:24 1:40:48 1:55:12 2:09:36 2:24:00 2:38:24 2:52:48

Time [Hours]

De

tect

or

rea

din

g [

V]

0א 0ב

0.3א 0.3ב

0.6א 0.6ב

1.25א 1.25ב

2.5א 2.5ב

5א 5ב

10א 10ב

NA concentration [PPM]

Detector reading for various NA concentration after about 2.5 hours

00.5

11.5

22.5

33.5

0 2 4 6 8 10 12

NA concentration [PPM]

Det

ecto

r re

adin

g [

V]

Reading a

Reading b

Graph 5– Measured voltage depend on time, for all NA concentration

Graph 6– Measured voltage depend on NA concentration, after 2.5 hours

Conclusions:

a. There is great importance to the bacteria OD (must be at least 0.2 OD). This is

probably the reason for the failure experiments in our first experiments.

b. The optimal time for sampling is about 1.5 hours (biggest difference in

voltage between different concentrations.

c. With respect to previous experiments ,in this experiment we didn't have a

decline for high NA concentration.

23

5. Dried Bacteria characterization:

Date: 29-31/7/09, 25-27/8/09, 1/9/09-3/9/09

Goal: find the differences between dried and liquid bacteria; prove the feasibility

of drying the bacteria in cuvettes.

Description:

First, we dried bacteria in small glass bottles (the orthodox way), as described in

the biological part of this document. We dried 3 bacteria concentrations (0.25,

0.35, 0.5 OD).

The technique for this is to grow and fresh the bacteria up to 0.25 OD, then

decrease its volume 20 times to 5 OD and by adding threalose9 we diluted the

bacteria and made it a suspension (2.5, 3.5, 5 OD).

We dried 4 bottles of each bacteria concentration (2 with copper stands in order to

improve heat conductance and 2 without it), each one contained 0.15 ml of

bacteria, 0.15 ml MOPS G (being condensed 10 times) and 0.015 ml PNPP (40

mg/ml concentration).

The drying process had taken 2 days, and then the dried bacteria were diluted with

1.5 ml of distilled water (in order to get 0.25, 0.35 and 0.5 OD).

We checked the reaction of the rehydrated bacteria by transferring 0.6 ml of it to 2

cuvettes. We add 24 cuvettes, for each bacteria concentration we had 2 cuvettes of

each NA concentration (0, 2.5 and 10 ppm).

After making the cuvettes we inserted it to an incubator and check the reaction

after 1.5 hours, 2.5 hours and 3 hours.

The next drying experiments were in order to check the drying process in

cuvettes. We had to find special stoppers and to use coupling for the cuvettes for

better heat conductance.

We dried the bacteria in the same way as in bottles (in each cuvette 0.06 ml of

bacteria, 0.06 ml of MOPS G condensed 10 times, 0.006 ml of PNPP of 40 mg/ml

concentration). We also used UV cuvettes in order to check their impact on drying

(they have different stoppers).

9 Substance used to facilitate the rehydration of the bacteria

24

After three hours

0

1

2

3

4

5

6

7

0 2 4 6 8 10 12

NA concentration [PPM]

De

tect

or

rea

din

g [

Vo

lt]

0.5 OD

0.35 OD

0.25 OD

After one hour and a half

0

2

4

6

8

10

12

14

16

0 2 4 6 8 10 12

NA concentration [PPM]

De

tect

or

rea

din

g [

Vo

lt]

0.5 OD

0.35 OD

0.25 OD

The drying process had taken 2 days, and then the dried bacteria were diluted with

0.6 ml of distilled water (in order to get 0.35 OD).

After making the cuvettes we inserted them into an incubator and checked the

reaction after 2 hours, 3 hours and 4.25 hours.

The last drying experiment was the same as the second one, only with 0.25 OD.

Results:

The results after 1.5 hours and 3 hours for the first experiment (drying in bottles)

are presented in graph 7 and graph 8.

Graph 7– Measured voltage vs. NA concentration, after 1.5 hours

Graph 8– Measured voltage vs. NA concentration, after 3 hours

25

The results after 3 hours for the second experiment (drying in cuvettes, OD –

0.35) are presented in graph 9.

With copper stands - after 3 hours

0

2

4

6

8

10

12

14

16

0 2 4 6 8 10 12

NA concentration [PPM]

De

tect

or

rea

qd

ing [

Vo

lt]

three hours

Graph 9– Measured voltage vs. NA concentration, after 3 hours

Conclusions:

a. We can see from graph 7 that the optimal OD for drying bacteria is 0.25 (same

as in liquid bacteria).

b. There is great importance during the drying process to heat conductivity (the

cuvettes are made of plastic) as well as to symmetry between the cuvettes.

c. There was a technical problem in the lyophilization machine, causing the

drying process to be unreliable.

d. For each set of drying bacteria, our system needs to be characterized.

6. Heating characterization – TEC vs. incubator:

Date: 31/7/09, 4/8/09.

Goal: determine the correct TEC temperature in order to get 37°C inside the

cuvette and comparing between the reaction inside an incubator and in a cuvette

coupled to a TEC.

Description: first we determine the temperature in the TEC in order to get to a

temperature of 37°C for the water inside the cuvette.

Then we took the bottles of dried bacteria, we transferred the rehydrated bacteria

to cuvettes with NA concentration of 2.5 ppm. We measured the reaction for

26

different bacteria concentration (0.25, 0.35, 0.5 OD) and put the cuvette inside the

incubator and similar cuvette coupled to a TEC.

We compared the measured voltage between the coupled to TEC cuvette and the

incubator inserted cuvette.

Results: in order to get 37°C for the water inside the cuvette, we need to set the

TEC up to 41°C.

The results for 0.25 OD is presented in table 1.

Time

[hours]

Incubator

[Volt]

TEC

[Volt]

1.5 3.47 3.61

2.5 2.16 2.17

Table 1 – 0.25 OD, 2.5 ppm NA concentration, comparing the measured voltage between TEC and incubator

Conclusions:

a. In order to get 37°C for the water inside the cuvette, we need to set the TEC

up to 41°C .

b. We see that there is no significant difference between TEC coupling

cuvettes and incubator inserted cuvettes.

7. Full system characterization:

Date: 8/9/09, 10/9/09, 17/9/09.

Goal: show the functionality of the system, determine the time and speed of the

pump, consider the steering issues and check the full integration of the system.

Description: We took the complete system and tested it on several NA

concentrations. We characterized the time and speed needed for 0.5L of water and

0.6ml of water to being pumped and checked the system reliability.

Results: as mentioned in the drying experiments, we had technical problem in the

lyophilization machine and therefore we had a problem to characterize the full

system reference voltage.

We made several sampling and saw the time needed for proper work of the

system and the parameters needed for the10.

10 For more details, see the system part of this document

27

Conclusions:

a. The system proof of concept is done – the system is working properly.

b. There is a need to characterize the reference voltage for each set of drying

bacteria and each toxin.

c. The system seems reliable, although it is needs to be improved when it comes

to puncturing the taps of the cuvettes and the reliability of the stack.

28

Chapter 5 – future work

The Bio-sensor system we presented in previous chapters has various options for

improvements and enhancements which we discovered during the ongoing work.

In this chapter we will present briefly the most important findings we have on the future

implementation of the system:

• Miniaturization and power saving: currently, the system is defined as a

prototype and it is based on several power supply and desktop computers. At

future deployment the system will have to be deployed in the field, which will

make it impossible to use desktops, and even notebooks might present some

problems. An optimum solution might be the NI CompactRio system.

CompactRio includes main processor and various other hardware modules such

as: Wireless module, Satellite communications and more. All the modules are in

rigid a case which is suitable for field deployment. One of the most important

benefits of this system is that the VI controlling the system can be easily exported

to the CompactRio, and save the development time.

Figure 10 -NI CompactRio system

• WSN – Wireless Sensor Networks is a relatively new discipline in the academy

and industry. It's main interest is creating sensor networks which :

o Share information between the sensors in minimum payload.

o Perform various calculations on the information shares.

o Change deployment and information sharing on the fly.

o Raise issues to the network manager.

Bio-sensor network which will work in WSN mode will enable a lot of vital

information that can enhance even further the system's capabilities. For example:

o Detecting malicious attacks on the Bio-sensor system.

29

o Monitor pollution propagating in water sources.

o Cross checking of results from a certain system.

A good example of WSN deployment is of systems that are on the shores of a

water source. These systems communicate with each other and send the

aggregated information to a base station.

• Detecting a large number of toxics: as presented in the previous chapters,

during the stage of genetically engineering the E.coli reporter strains, it is possible

to design strains that will be activated by a different spectrum of toxic substances,

more over, it is possible that each strain will have a different reporting gene which

would activate different substrates, so the signal produced will be unique in

wavelength and/or intensity. We can use these options to discover a few toxins in

one system in one of the following methods :

o Create a few cuvettes in one system; each one contains a different type of

bacteria. In this way, in every sample we can discover a different type of

toxin.

o Create a cuvette which contains few bacteria, each one can sense a

different toxin. The reaction for each kind of toxin will change the

substrate color to a different one. Using advanced technique, it is possible

to discover which toxin or combination of toxins caused the specific

reactions.

• Complex data analysis - The system prototype can only discover the existence of

a toxin or lack of it, as described in the previous chapters. More complex data

analysis that includes performing various mathematics calculations on the data

from the system might obtain the level of toxicity as well. For this analysis to take

place there is a need to characterize further the reaction of the bacteria to all level

of toxins.

• Enabling continuous sampling – As for now, the system is limited to 10

cuvettes, which enables limited number of samples. There is a need to allow

more cuvettes to be part of the system, and yet allowing easy replacements of the

cuvette stack.

30

• Adding controls: - As for now, the system lacks internal controls. The purpose of

the controls is to know that the system is working properly. There are 3 different

controls :

o Positive controls: In these controls, we will use cuvettes with a known

amount of toxin, and add clean water, not from the water source. The

purpose of this control is to know that our system is working properly, and

discover the known amount of toxin.

o Negative control: In this control, we will use cuvettes with bacteria and a

clean water sample. This control will help us to calibrate the system and

know we don’t get false alarms.

o Constructive control – in this control, we will use cuvettes with a know

amount of toxin, and water from the water source. If we will get a low

read in this control as well as in a sample from the reservoir, this means

that the water sample has a cytotoxic effect on the bacteria and that the

bacteria died before reacting with the substrate. If we get high reading in

the control this means that the water in the water source had little amount

of toxin, plus the toxin we added and the reading is reliable.

In order to add this control, some changes will have to be done in the controlling

VI. As part of the prototype, we inserted some basic infrastructure to enable easy

addition of the control. Also, the system should have access to a clean water

source or alternatively, add clean water to the cuvettes in an enclosed cell. When

the needle goes down on these special cuvettes, then this cell will be breached and

clean water will get into the cuvette.

Memebrane

Clean water cell

Needle arrive into the

cuvette and allows

clean water to flow to

the cuvette

Figure 11 -Clean water cell sample

The needle arrives inside the cuvette and allows clear water to flow into it

31

Summary We managed to characterize and build a system capable of detecting pollutants in water

in a fast, automated and simple manner. Our system makes use of bacteria producing a

colorimetric reaction never before used in a biosensor which includes a complete

sampling and detecting system. The use of such bacteria enables minimizing the time

needed to reach a conclusion about the pollution in comparison with other types of

bacteria usually used in such systems – mainly the fluorescent bacteria.

While the ways of detecting pollutant in water reservoirs used today take up to 24 hours

and require reaching the reservoir whenever a sample is taken, our system enables fast

detection and monitoring from a distance with the need for maintenance only once every

few months, thanks to the use of freeze-dried bacteria.

We present a prototype showing the feasibility of such an all included biosensor using

colorimetric bacteria. In order to improve the system in the future one would have to

examine ways of minimization and cost lowering as well as conduct further biological

experiments and better process the bacteria's reaction to enable more complex data to be

extracted from it. The drying process is yet to mature and is still under investigation in

several groups around the world. It needs to be fitted specifically to a system such as ours

in order to make the best of it when taking our need under consideration.

32

References 1. Ali Niazi and Ateesa Yazdanipour - Spectrophotometric simultaneous determination of

nitrophenol isomers by orthogonal signal correction and partial least squares 2. Arthur Rabner MA thesis – TAU.

3. Baselt, D.R., Lee, G.U., Hansen, K.M., Chrisey, L.A., Colton, R.J. (1997) A high-

sensitivity micromachined biosensor. Proc. IEEE 85: 672-680.–

4. Belkin, S.(2003) Microbial whole-cell sensing systems of environmental pollutants Curr.

Opin. Microbiol. 6: 206–212.

5. Kim, B.C., Gu, M.B. (2005) A Multi-Channel Continuous Water Toxicity Monitoring

System: Its Evaluation and Application to Water Discharged from a Power Plant.

Environ. Monit. Assess. 109:,123-133

6. Frederick C. Neidhardt, Philip L. Bloch and David F. Smith - Culture Medium for

Enterobacteria J.bacteriology 1974 119/3 p.736-747

7. A luxCDABE-based bioluminescent bioreporter for detection of phenol – Abd-El-

Haleem D, Ripp S, Scott C, Sayler GS - J Ind Microbiol Biotechnol. 2002

Nov:29(5):233-7.

8. Towards toxicity detection using a lab-on-chip based on the integration of MOEMS and

whole-cell sensors - Noel M. Elman, Hadar Ben-Yoav , Marek Sternheim, Rachel Rosen,

Slava Krylov, Yosi Shacham-Diamand

9. Recombinant Bacterial Reporter Systems - Shimshon Belkin

10. A Dual-Color Bacterial Reporter Strain for the Detection of Toxic and Genotoxic Effects

- N. Hever, S. Belkin

11. Bioreporters: gfp versus lux revisited and single-cell response - Stefanie Kohlmeier,

Matthew Mancuso, Robin Tecon, Hauke Harms, Jan Roelof van der Meer, Mona Wells.

12. Fluorescence and bioluminescence reporter functions in genetically modified bacterial

sensor strains - Eran Sagi, Navit Hever, Rachel Rosen, Amelita J. Bartolome, J. Rajan

Premkumar, Roland Ulber, Ovadia Lev, Thomas Scheper, Shimshon Belkin.

13. http://www.nuncbrand.com/NAG/DP0010.htm

14. Biran, A. (2009) Genetically engineered bacteria for electrochemical detection of

genotoxicants. MSc Thesis. Plant and Environmental Sciences, the Alexander Silberman

Institute for Life Sciences, Faculty of Science, the Hebrew University of Jerusalem, Israel

33

Appendix 1 – Data sheets Data sheet of the LED used in our system.

From: http://docs-asia.origin.electrocomponents.com/webdocs/02a0/0900766b802a04e3.pdf

34

Data sheet of the pump used in our system

From: http://mrceng2.b-smart.co.il/Media/Uploads/BT00300TSPEC(1).pdf

35

Data sheet of the photodiode used in the detector in our system.

36

From: http://jp.hamamatsu.com/resources/products/ssd/pdf/s1223_series_kpin1050e01.pdf

37

Appendix 2 – Labview control program

And inside the stacked sequence:

38

39

40

Appendix 3 – NI CompactRio Spec

41

42

Appendix 4: Minimal medium MOPS glucose MOPSX10 – after preparing and filtering with 0.22µm filter, it is preferred to divide it to 50ml sterile tubes and to keep it in -20OC. Mixing the following solutions in the given order Conc. MW ml g coments

MOPS PH=8.2 with KOH

1M 209.26 400 83.7 Freshly prepared

Tricine pH=8.2 with KOH

1M 179.2 40 7.168 Freshly prepared

FeSO4 0.01M 278 10 0.0278 Freshly prepared

NH4Cl 1.9M 53.49 50 5.1

K2SO4 0.276M 174.27 10 0.48

CaCl2 5*10^-4M

111 10 7.35mg/100ml

MgCl2*6H2O 0.528M 203.3 10 1.073424 NaCl 5M 58.44 100 29.22 micronutrients 10 H2O 360

Total 1000 You prepare the micronutrients separately and add 10 µl to the medium. You can keep it after filtration in 4°c.

Micronutrients solution Conc. MW

(NH4)6(Mo7)24 3*10-6M

H3BO3 4*10-4M 61.83

CoCl2 3*10-5 237.9

CuSO4 10-5 249.68

MnCl2 8*10-5 161.88

ZnSO4 10-5 287.54

MOPS glucose+0.1% Yeast extract- 0.5l MOPS*10 50ml 0.0132M K2HPO4 5 ml D-glucose 1g YE 0.5g Total 500ml (445ml H2O) After the medium is ready, filtration with 0.22µm mc.

43

Appendix 5: an experiment protocol . rfaE/2TTSsulA::phoA∆של O/Nתרבית .1

.OD=0.25גידול שעתיים עד ). כלי זכוכית(בארלנמייר , MOPS - ב 1:100רענון .2 ):קיווטות יש ליד הספקטרופוטומטר(הכנת סטוקים וקיווטות , בזמן הרענון .3אליהם מוסיפים , mg pNPP 8שוקלים ( mg/ml 0.8 - בריכוז MOPS -מומס ב pNPPסטוק -

10ml MOPS( ,צלזיוס 20- בטמפרטורה של , הסובסטראט נמצא במקפיא .לכל קיווטה MOPS+pNPP( ,300µl(מחלקים את המדיום שכבר מכיל את הסובסטראט - . במבחנות אפנדורף )Nalidixic acid )NAמכינים סטוק שריכוז הרצוי של המשרן -

1000ppm NA= 90µl DDW+10µl NA 10,000ppmהכנת סטוק של • 100ppm NA= 90µl DDW+10µl NA 1000 ppmק של הכנת סטו •מכל ריכוז מכינים דופליקאט - הוספת הנפח הרצוי לכל אחת מהקווטות מהסטוק הנכון •

ניכנס לתוך נפח הנוזל בגלל הנפחים הקטנים NA -חשוב לוודא שהטיפ המכיל את ה( ...)המוספים

נפח הוספה ריכוז סופי בקיווטה סטוק1000 ppm 10 ppm 6 µl 1000 ppm 5 ppm 3 µl 1000 ppm 2.5 ppm 1.5 µl 100 ppm 1.25 ppm 7.5 µl 100 ppm 0.6 ppm 3.75 µl 100 ppm 0.3 ppm 1.9 µl

0

חיידקים מרועננים µl 300מוסיפים לכל קיווטה , לאחר הכנת כל הקיווטות עם הריכוזים הנכונים - . הרצוי OD -שהגיעו ל

)'דק 60 - סטנדרטי(ומכניסים לאינקובאטור לזמן הרצוי , ם כיסוי פלאטותסוגרים את הקיווטות ע -