Difference between revisions of "Team:Queens Canada/Overview"

Line 4: Line 4:
 
<html>
 
<html>
 
<head>
 
<head>
−
<link rel="stylesheet" href="https://2014.igem.org/Team:Imperial/Resources/CSS:pure-nr?action=raw&ctype=text/css">
+
<base target="_self">
−
<link rel="stylesheet" href="https://2014.igem.org/Team:Imperial/Resources/CSS:pure-skin?action=raw&ctype=text/css">
+
 
<style>
 
<style>
  
−
#locator1 {
+
p.small {
−
position:absolute;
+
line-height: 70%;
−
left:30px;
+
−
top:540px;
+
−
z-index:200;
+
−
 
+
 
}
 
}
  
−
#locator2 {
+
p.big {
−
position:absolute;
+
line-height: 200%;
−
left:90px;
+
−
top:720px;
+
−
z-index:200;
+
 
}
 
}
  
−
#locator3 {
+
div.lightBlue {
−
position:absolute;
+
padding: 0px 110px;
−
left:700px;
+
background-color: #3399ff;
−
top:600px;
+
−
z-index:200;
+
 
}
 
}
  
−
#locator4 {
+
ul {
−
position:absolute;
+
font face: Corbel;
−
left:40px;
+
line-height: 1.5;
−
top:840px;
+
font size: 3;
−
z-index:200;
+
color: black;
 
}
 
}
  
−
#locator5 {
+
div#afixRight{
−
position:absolute;
+
width:70px;
−
left:670px;
+
position:fixed;
−
top:900px;
+
bottom:12%;
−
z-index:200;
+
right:0;
 +
z-index:11;
 
}
 
}
  
−
#locator6 {
+
div.center {
−
position:absolute;
+
padding: 10px 10px;
−
left:70px;
+
align: center;
−
top:1030px;
+
−
z-index:200;
+
−
}
+
−
 
+
−
#locator7 {
+
−
position:absolute;
+
−
left:150px;
+
−
top:1200px;
+
−
z-index:200;
+
 
}
 
}
  
 
p {
 
p {
−
color: white;
+
color: black;
−
}
+
−
 
+
−
li {
+
−
color: white;
+
 
}
 
}
−
 
  
 
</style>
 
</style>
 
</head>
 
</head>
 +
<body>
  
−
<div class="column full_size" >
+
<div>
−
<img src="https://static.igem.org/mediawiki/2017/5/5e/T--Queens_Canada--OverviewBanner.jpeg">
+
  <img src="https://static.igem.org/mediawiki/2017/8/82/T--Queens_Canada--InterlabStudyBanner.jpg" style="width:100%">
−
</div><br>
+
</div>  
  
 +
<div style="height:30px;width:100%">
 +
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px">
 +
</div>
 +
 +
<div style="background-color:#afbee8;color:white;padding:50px 150px 50px 150px;">
 +
<center>
 +
<p><font size="5" color="black">Queen's Canada iGEM team is very excited to be a part of the 2017 Interlab study. This study builds upon previous years attempts to answer one major question: <br><br> How close can the numbers be when fluorescence is measured all around the world?</font></p></center>
 
<br>
 
<br>
−
<br>
+
</div>
−
<br>
+
  
 +
<div style="height:30px;width:100%">
 +
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px">
 +
</div>
  
−
<div id="overview">
+
<div style="background-color:#afbee8;color:white;padding:5px 50px 50px 50px;line-height:150%;">
−
<br>
+
<img class="bioleft" align="left" src="https://static.igem.org/mediawiki/2017/c/cf/T--Queens_Canada--Fluorescence.png" height="auto" width="35%" style="object-fit:contain;padding:50px 50px 10px 10px;">
−
<h1 style="color: red">Overview</h1>
+
<font face="Corbel" font style="line-height:1.5" font size="2" color="black">
−
<br>
+
 
−
<p>The Arctic is predicted to contain approximately 22% of the world’s oil and natural gas resources locked up beneath its basins. Due to the lack of natural resources in the surrounding tundra biome, bioremediation of oil spills can be especially challenging in the Arctic. As a result, an effective and environmentally safe method of petroleum cleanup may come in high demand during a major oil spill, which has been shown to incur costs ranging from the hundred millions to billions of dollars for conventional containment and cleanup. The QGEM Team is turning to nature as our inspiration for building a safer and cheaper method of oil spill cleanup, using synthetic biology as our tool. This summer, we will be designing and engineering a bacterial biofilm-based material that functions to bind ice and recruit oil-degrading native marine bacteria. This will be done by engineering a protein naturally expressed in bacterial biofilm, CsgA, to append an ice-binding protein and a bacterial adhesin domain. The end product will be a dynamic, bifunctional biomaterial that may be deployed into the Arctic marine environment during oil spills as a bioremediation factory, with limited disturbance to the surrounding ecosystem. </p>
+
<Div style="position: absolute; top:1610px; left:140px; width:400px; height:25px">Agar plate streaked with negative control device (left) and positive control<br> device (right) viewed under UV light. Looks like they work!</Div>
−
<br>
+
 
 +
 
 +
<p class="big"><font size="5" color="black"><br><br><br><br>In 2017, the Queen's University iGEM team is participating in the Fourth International Interlaboratory Measurement Study for their first time. The Interlab Study aims to analyze and improve the replicability of fluorescence measurements. This year, the reliability of some RBS devices (BCDs) will be tested all around the world, using the expression of green fluorescent protein (GFP) to quantify translation.<br><br> Participating iGEM teams measure fluorescence exhibited by the GFP across six test devices, each with different RBS sequences. A positive and negative control are also used to calculate expression levels using fluorescence/OD600. Reproducibility is one of the most challenging and critical aspects of scientific research. We hope that the data from this study can help establish a baseline for GFP measurement reproducibility, given GFP's widespread use as a tool in molecular biology. This collaborative effort is a small but meaningful step towards this goal.</font></p>
 
<br>
 
<br>
 +
<br><br><br><br><br><br><br><br>
 
</div>
 
</div>
  
  
−
<div id="biofilms">
+
<div style="height:30px;width:100%">
−
<br>
+
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px">
−
<h1 style="color: red">Biofilms</h1>
+
−
<p><font size="3" face="Lucida Sans Unicode">In nature, the majority of bacteria exist as biofilms. Biofilms are organized bacterial communities that grow on an extracellular matrix scaffold composed mostly of polysaccharides, proteins and nucleic acids [2]. Biofilms typically carry negative connotations, particularly in medical settings where they are associated with antibiotic-resistant infections. However, we can exploit the same traits that make biofilms a formidable healthcare challenge to create engineered biomaterials. Bacteria in biofilms have several characteristic advantages over their planktonic (free-living) counterparts [2]:
+
−
<ul>
+
−
<li>Enhanced resource capture and processing</li>
+
−
<li>Facilitated social communication (quorum sensing)</li>
+
−
<li>Protection against environmental stress (i.e. antibiotics, shear stress)</li>
+
−
</ul></p>
+
−
<br>
+
−
<p>Our engineered <i>E. coli</i> (expressing our CsgA fusions) can provide the oil-degrading <i>M. hydrocarbonoclasticus</i> with an ice-binding biofilm scaffold, significantly increasing its ability to survive and degrade hydrocarbons in harsh Arctic conditions.  
+
−
</font></p>
+
 
</div>
 
</div>
−
<br>
 
−
<br>
 
  
−
<div id="csga">
+
 
−
<br>
+
 
−
<h1 style="color: red">CsgA + curli</h1>
+
<div style="background-color:#afbee8;color:white;padding:50px 150px 50px 150px;">
−
<br>
+
 
−
<p>CsgA is an amyloid protein monomer that polymerizes to form long nanofibres. Assembled CsgA nanofibres are referred to as curli. CsgA accounts for the majority of the proteinaceous component of E. coli biofilms. CsgA is secreted from the cells, and in wild-type E. coli it self-assembles to form curli nanofibres tethered to the cell surface by the nucleator protein CsgB. CsgA can also polymerize without CsgB in vitro. CsgA is relatively tolerant of C-terminal peptide fusions, granted they are less than about 40 amino acids in length. As an amyloid, CsgA is rich in β-sheet structure and remarkably resistant to denaturing conditions.</p>
+
 
−
<br>
+
<h1><font color="black" face="Arial"><center><span style="font-weight:normal; font-size: 23pt">Background</span></center></font></h1><hr/>
 
<br>
 
<br>
 +
<p class="big"><font size="5" face="Corbel"> The ability to reproduce results in biological systems is difficult due to the stochastic nature of living cells and inconsistent laboratory practices [1]. Comparing quantitative results between experiments is often difficult with many variables impacting the results. These may include:</font></p>
 +
<ul><font size="4" face="Lucida Sans Unicode"><li>The various instruments used and their different calibrations</li>
 +
<li>Variation in laboratory practices/protocols</li>
 +
<li>Systematic variability e.g. differences in strains used, physical laboratory conditions</li>
 +
<li>Variation in interpreting and communicating results</li>
 +
</font>
 +
</ul>
 +
</div>
 +
 +
<div style="height:30px;width:100%">
 +
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px">
 
</div>
 
</div>
  
−
<div id="afp8">
+
 
 +
<div style="background-color:#afbee8;color:white;padding:50px 150px 50px 150px;">
 +
 
 +
<h1><font color="black" face="Arial"><center><span style="font-weight:normal; font-size: 23pt">Methods and Materials</span></center></font></h1><hr/>
 
<br>
 
<br>
−
<h1 style="color: red">AFP8</h1>
 
 
<br>
 
<br>
−
<p>Antifreeze proteins are a subset of ice-binding proteins. They adhere to the crystalline structure of ice by organizing ice-like water molecules on their surface. This makes further ice growth thermodynamically unfavourable. Although they vary greatly in structure, the ice-binding sites all have conserved properties. Here, we used the 37-residue, α-helical Type I AFP (we call it AFP8) from the winter flounder fish. We fused AFP8 directly to the C-terminus of CsgA. Most organisms that use AFPs circulate them in their blood as protection from freezing temperatures. However, AFPs can also be used to adhere to solid ice. In fact, the Arctic marine bacterium Marinomonas prymoriensis successfully uses an ice-binding domain as part of its large adhesin complex to keep itself at the top of the water column, where oxygen and nutrients are abundant [7].</p>
+
 
 +
<div class="center"><img src="https://static.igem.org/mediawiki/2017/2/21/T--Queens_Canada--Protocol.png" alt="Interlab study protocols." height ="auto" width=70%"></div>
 +
<font face="Corbel" font style="line-height:1.5" font size="2" color="black"><figcaption><center>Interlab Study Protocols</font>
 
<br>
 
<br>
 
<br>
 
<br>
−
</div>
 
  
−
<div id="marinobacter">
+
 
 +
<p class="big"><font size="5" color="black" face="Corbel"><We transformed the six plasmids containing the six test device constructs (J364000, J364001, J364002, J364003, J364004, J364005) as well as the positive and negative controls into the <i>E. coli</i> DH5a strain. After picking two colonies from each transformation, we grew up the cells and started the calibration protocols of OD600 reference point using the LUDOX solution. FITC was used as the standard for fluorescence. We used a Molecular Devices SpectraMax M2e plate reader on the topreading setting for both our OD600 and fluorescent measurements. Black 96-well plates with clear bottoms were used. We measured fluorescence at an excitation wavelength 395nm and emission wavelength of 508nm [2].</font></p>
 
<br>
 
<br>
−
<h1 style="color: red"><i>Marinobacter hydrocarbonoclasticus</i></h1>
 
 
<br>
 
<br>
−
<p><font size="3" face="Lucida Sans Unicode">As the name of this Gram-negative bacteria would suggest, it loves to degrade petroleum hydrocarbons. The species was first isolated from the waters of the Mediterranean, where it frequently forms oleolytic biofilms [3, 4]. The hydrophobic organic molecules that it degrades are used as a source of carbon and metabolic energy. We had originally considered directly fusing hydrocarbon-degrading enzymes (i.e. ethylbenzene dehydrogenase) onto CsgA. However, since individual enzymes function poorly outside of the cell, we opted to append whole <i>M. hydrocarbonoclasticus</i> bacterial cells instead. When it comes to optimal hydrocarbon degradation, you can’t beat Nature!</font></p>
+
 
 +
<font face="Arial" font size="5"><a href="https://static.igem.org/mediawiki/2017/8/85/InterLab_2017_Plate_Reader_Protocol.pdf" style = "color:rgb(128,128,128); text-decoration:none;"><font color="#808080">Interlab Protocols (PDF)</font></a>
 +
 
 
<br>
 
<br>
 
<br>
 
<br>
 +
</font>
 
</div>
 
</div>
  
−
<div id="spyTag">
+
 
 +
<div style="height:30px;width:100%">
 +
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px;">
 +
</div>
 +
 
 +
<div style="background-color:#afbee8;color:white;padding:50px 150px 50px 150px;"
 +
 
 +
<h1><font color="black" face="Arial"><center><span style="font-weight:normal; font-size: 23pt">
 +
Results and Discussion
 +
</span></center></font></h1><hr/>
 +
 
 
<br>
 
<br>
−
<h1 style="color: red">SpyTag/SpyCatcher</h1>
 
 
<br>
 
<br>
−
<p><font size="3" face="Lucida Sans Unicode">SpyTag and SpyCatcher form two halves of a split protein system. When these two domains come together, they spontaneously form a new covalent peptide bond. This permanently links SpyTag and SpyCatcher, along with whatever was attached to them. We used this system to overcome CsgA export size limitations. Nguyen et al. found that CsgA with fusion peptides larger than approximately 40 amino acids were too large to be properly exported [5]. However, SpyTag is only 13 amino acids long, so we used CsgA-SpyTag fusions to enable us to add larger domains than could otherwise be exported. Our Lectin-SpyCatcher fusion would then be added to the exported CsgA-SpyTag, permanently appending our large (30 kDa) Lectin to CsgA.
+
<div class="center">
−
</font></p>
+
<img src="https://static.igem.org/mediawiki/2017/b/b0/T--Queens_Canada--Fluorescein.png" alt="absorbance at 600 nm" height="auto" width="50%">
 +
<font face="Corbel" font style="line-height:1.5" font size="2" color="black"><figcaption><center><b>Fig 1.</b> Fluorescein standard curve.</font></center></figcaption></div>
 
<br>
 
<br>
 +
 +
<p class="big"><font size="5" color="black" face="Corbel">The above fluorescence calibration curve (Fig. 1) was created by measuring fluorescence intensity of different concentrations of fluorescein.</font></p>
 +
 +
<div class="center">
 +
<img src="https://static.igem.org/mediawiki/2017/9/9c/T--Queens_Canada--FI_Graph_Normalized_Interlab_Study.png" alt="Fluorescents of the test devices." height="auto" width="50%">
 +
<font face="Corbel" font style="line-height:1.5" font size="2" color="#b0bee1"><figcaption><center><b>Fig 2.</b> This graph shows the fluorescence of each test device measured every two hours over a span of 6 hours.</font></center></figcaption></div>
 
<br>
 
<br>
−
</div>
 
  
−
<div id="sugar">
+
<p class="big"><font size="5" color="black" face="Corbel">Cultures were sampled at the += 0,2,4,6 hour marks in 500ml aliquots from 10ml cultures. All samples were added to a 96-well plate to measure fluorescence intensity. Fluorescent values were normalized prior to plotting. All points on the graphs are the average of the two colonies grown (which themselves are the average of 4 wells each). Test device 2 had the greatest overall increase in fluorescence. Both the negative control device and the LB+ chloramphenicol sample had no significant increase in fluorescence.</font></p>
 +
 
 
<br>
 
<br>
−
<h1 style="color: red">Sugar binding</h1>
+
<div class="center">
 +
<img src="https://static.igem.org/mediawiki/2017/e/e1/T--Queens_Canada--OD600.png" alt="absorbance at 600 nm" height="auto" width="50%">
 +
<font face="Corbel" font style="line-height:1.5" font size="2" color="#b0bee1"><figcaption><center><b>Fig 3.</b> This graph shows the absorbance at 600 nanometres of each cell culture, which<br>provides an estimate for the number of cells in the samples.</font></center></figcaption></div>
 +
 
 +
<p class="big"><font size="5" color="black" face="Corbel">Every test device exhibits steady, somewhat sigmoidal bacterial growth. The LB + chloramphenicol sample shows no change in OD600 over time.</font></p>
 +
</div>
 +
 
 +
<div style="height:30px;width:100%">
 +
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px">
 +
</div>
 +
 
 +
<div style="background-color:#afbee8;color:white;padding:50px 150px 50px 150px;">
 +
 
 +
<h1><font color="black" face="Arial"><center><span style="font-weight:normal; font-size: 23pt">Conclusions</span></center></font></h1><hr/>
 
<br>
 
<br>
−
<p>Sugar binding domains are used to tether the hydrocarbon–degrading M. hydrocarbonoclasticus (MH) to the CsgA scaffold via dextran. Not enough is known about MH physiology to make genetic modification (i.e. expression of SpyCatcher) of this exotic organism practical. Thus, we made use of an existing MH surface protein: a sugar binding domain. The sugar binding domain is a part of Region III of the large MhLap adhesin complex expressed on the surface of MH. Our affinity chromatography and microfluidics experiments proved this domain binds dextran. A lectin domain was appended to CsgA to enable the curli biofilm to latch onto dextran as well. Dextran thus serves to crosslink MH to the E. coli CsgA biofilm. We chose the C-type lectin from residues 403-532 of the human polycystin-1 protein, encoded for by the PKD1 gene [6]. Lectins bind sugar moieties of many polysaccharides, playing key roles in cellular recognition in many organisms. </p>
+
<font face="Corbel" font style="line-height:1.5" font size="5" color="black">
 +
<ul><li>The Queen's_Canada iGEM team was grateful for the opportunity to contribute to the Interlab Study for the first time.</li>
 +
<li>It appears that our cells only began expressing significant amounts of GFP after the 4-hour mark. One would expect the curve of increasing GFP fluorescence to mirror the curve of OD600, if GFP expression is truly constitutive. The OD600 curve shows steady, somewhat sigmoidal growth, while the fluorescent intensity curve is a plateau until after 4 hours have elapsed. </li>
 +
<li>This suggests either a certain threshold concentration of GFP is required to be detectable by our plate reader, or that GFP expression only begins at a certain cell density threshold (which is reached at approximately 0.2 OD on our Figure 3 graph). </li>
 +
<li> Both the LB + chloramphenicol and negative control wells showed no significant increase in fluorescence, as expected. <li?>
 +
</font>
 +
 
 
</div>
 
</div>
  
−
<br>
+
<img src="https://static.igem.org/mediawiki/2017/6/64/T--Queens_Canada--NavyNavBackground1.jpg" style="width:100%;height:30px">
−
<br><br>
+
</div>
−
<br><br>
+
 
−
<br><br>
+
<div style="background-color:#afbee8;color:white;padding:50px 150px 50px 150px;">
−
<br><br>
+
 
−
<br><br>
+
<h1><font color="black" face="Arial"><center><span style="font-weight:normal; font-size: 23pt">References</span></center></font></h1><hr/>
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
−
<br>
+
−
<br><br>
+
−
<br><br>
+
−
<br><br>
+
 
<br>
 
<br>
  
 +
<ol><font style="line-height:1.5" "left-margin:0" font face="corbel" font size="2" color="black">
 +
<li>Kwok, R. 2010. Five hard truths for synthetic biology. Nature, 463, 288.</li>
 +
<li>Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., Prasher, D. C. 1994. Green Fluorescent Protein as a Marker for Gene Expression. Science: 263(5148), 802-805. </li>
 +
</font></ol>
 +
 +
</body>
  
 
</html>
 
</html>
 +
<br>
 +
<br>
 +
<br>

Revision as of 03:07, 30 October 2017

Queen's Canada iGEM team is very excited to be a part of the 2017 Interlab study. This study builds upon previous years attempts to answer one major question:

How close can the numbers be when fluorescence is measured all around the world?


Agar plate streaked with negative control device (left) and positive control
device (right) viewed under UV light. Looks like they work!





In 2017, the Queen's University iGEM team is participating in the Fourth International Interlaboratory Measurement Study for their first time. The Interlab Study aims to analyze and improve the replicability of fluorescence measurements. This year, the reliability of some RBS devices (BCDs) will be tested all around the world, using the expression of green fluorescent protein (GFP) to quantify translation.

Participating iGEM teams measure fluorescence exhibited by the GFP across six test devices, each with different RBS sequences. A positive and negative control are also used to calculate expression levels using fluorescence/OD600. Reproducibility is one of the most challenging and critical aspects of scientific research. We hope that the data from this study can help establish a baseline for GFP measurement reproducibility, given GFP's widespread use as a tool in molecular biology. This collaborative effort is a small but meaningful step towards this goal.










Background


The ability to reproduce results in biological systems is difficult due to the stochastic nature of living cells and inconsistent laboratory practices [1]. Comparing quantitative results between experiments is often difficult with many variables impacting the results. These may include:

  • The various instruments used and their different calibrations
  • Variation in laboratory practices/protocols
  • Systematic variability e.g. differences in strains used, physical laboratory conditions
  • Variation in interpreting and communicating results
    •

Methods and Materials



Interlab study protocols.
Interlab Study Protocols

E. coli DH5a strain. After picking two colonies from each transformation, we grew up the cells and started the calibration protocols of OD600 reference point using the LUDOX solution. FITC was used as the standard for fluorescence. We used a Molecular Devices SpectraMax M2e plate reader on the topreading setting for both our OD600 and fluorescent measurements. Black 96-well plates with clear bottoms were used. We measured fluorescence at an excitation wavelength 395nm and emission wavelength of 508nm [2].



Interlab Protocols (PDF)

Results and Discussion


absorbance at 600 nm
Fig 1. Fluorescein standard curve.

The above fluorescence calibration curve (Fig. 1) was created by measuring fluorescence intensity of different concentrations of fluorescein.

Fluorescents of the test devices.
Fig 2. This graph shows the fluorescence of each test device measured every two hours over a span of 6 hours.

Cultures were sampled at the += 0,2,4,6 hour marks in 500ml aliquots from 10ml cultures. All samples were added to a 96-well plate to measure fluorescence intensity. Fluorescent values were normalized prior to plotting. All points on the graphs are the average of the two colonies grown (which themselves are the average of 4 wells each). Test device 2 had the greatest overall increase in fluorescence. Both the negative control device and the LB+ chloramphenicol sample had no significant increase in fluorescence.


absorbance at 600 nm
Fig 3. This graph shows the absorbance at 600 nanometres of each cell culture, which
provides an estimate for the number of cells in the samples.

Every test device exhibits steady, somewhat sigmoidal bacterial growth. The LB + chloramphenicol sample shows no change in OD600 over time.

Conclusions


  • The Queen's_Canada iGEM team was grateful for the opportunity to contribute to the Interlab Study for the first time.
  • It appears that our cells only began expressing significant amounts of GFP after the 4-hour mark. One would expect the curve of increasing GFP fluorescence to mirror the curve of OD600, if GFP expression is truly constitutive. The OD600 curve shows steady, somewhat sigmoidal growth, while the fluorescent intensity curve is a plateau until after 4 hours have elapsed.
  • This suggests either a certain threshold concentration of GFP is required to be detectable by our plate reader, or that GFP expression only begins at a certain cell density threshold (which is reached at approximately 0.2 OD on our Figure 3 graph).
  • Both the LB + chloramphenicol and negative control wells showed no significant increase in fluorescence, as expected.

References


  1. Kwok, R. 2010. Five hard truths for synthetic biology. Nature, 463, 288.
  2. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., Prasher, D. C. 1994. Green Fluorescent Protein as a Marker for Gene Expression. Science: 263(5148), 802-805.