Limits...
Quantification of mRNA in single cells and modelling of RT-qPCR induced noise.

Bengtsson M, Hemberg M, Rorsman P, Ståhlberg A - BMC Mol. Biol. (2008)

Bottom Line: The noise is insignificant for initial copy numbers >100 while at lower copy numbers the noise intrinsic of the PCR increases sharply, eventually obscuring quantitative measurements.Noise in single-cell RT-qPCR is insignificant compared to biological cell-to-cell variation in mRNA levels for medium and high abundance transcripts.To minimize the technical noise in single-cell RT-qPCR, the mRNA should be analyzed with a single RT reaction, and a single qPCR reaction per gene.

View Article: PubMed Central - HTML - PubMed

Affiliation: Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, The Churchill Hospital, Oxford, OX3 7LJ, UK. martin.bengtsson@med.lu.se

ABSTRACT

Background: Gene expression has a strong stochastic element resulting in highly variable mRNA levels between individual cells, even in a seemingly homogeneous cell population. Access to fundamental information about cellular mechanisms, such as correlated gene expression, motivates measurements of multiple genes in individual cells. Quantitative reverse transcription PCR (RT-qPCR) is the most accessible method which provides sufficiently accurate measurements of mRNA in single cells.

Results: Low concentration of guanidine thiocyanate was used to fully lyse single pancreatic beta-cells followed by RT-qPCR without the need for purification. The accuracy of the measurements was determined by a quantitative noise-model of the reverse transcription and PCR. The noise is insignificant for initial copy numbers >100 while at lower copy numbers the noise intrinsic of the PCR increases sharply, eventually obscuring quantitative measurements. Importantly, the model allows us to determine the RT efficiency without using artificial RNA as a standard. The experimental setup was applied on single endocrine cells, where the technical and biological noise levels were determined.

Conclusion: Noise in single-cell RT-qPCR is insignificant compared to biological cell-to-cell variation in mRNA levels for medium and high abundance transcripts. To minimize the technical noise in single-cell RT-qPCR, the mRNA should be analyzed with a single RT reaction, and a single qPCR reaction per gene.

Show MeSH

Related in: MedlinePlus

Noise levels for the measurements of known quantities of mRNA from Rps29, Chgb, Ins2, Gfap, Nes and Sox2. (a) The measured noise strength (η2 = SD2/mean2) for Rps29 (blue circles), Chgb (red squares) Ins2 (black triangles), Gfap (green triangles), Nes (cyan stars) and Sox2 (magenta diamonds) with corresponding fits (lines) obtained using non-linear regression for the mathematical model presented in Additional file 2. The model estimates the RT-efficiency and the results are reasonable for five of the six genes. For Ins2 the efficiency is much lower than expected. The poor curve fit for the Ins2 gene results from the fact that the Ins2 data was generated for much higher copy numbers whereas our model for the PCR-noise was adapted for the low abundances of tge five other genes. (b) The proportion of the total noise for the PCR (filled symbols) and RT reactions (open symbols). Circles, squares and triangles are designated as in (a). The PCR and RT components do not have to add up to 1; the noise stemming from the dilution corresponds to the remaining noise. For Rps29, Chgb and Gfap, the PCR noise clearly dominates for all concentrations. For Ins2 and Sox2, the estimated RT efficiency is very low which means that this reaction will add a larger contribution to the total noise. Nes has an intermediate efficiency and for low copy numbers the RT noise dominates but it becomes smaller than the qPCR nosie when more transcripts are analyzed. Furthermore, the copy numbers are relatively high for Ins2 which deflates the PCR noise.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2483285&req=5

Figure 3: Noise levels for the measurements of known quantities of mRNA from Rps29, Chgb, Ins2, Gfap, Nes and Sox2. (a) The measured noise strength (η2 = SD2/mean2) for Rps29 (blue circles), Chgb (red squares) Ins2 (black triangles), Gfap (green triangles), Nes (cyan stars) and Sox2 (magenta diamonds) with corresponding fits (lines) obtained using non-linear regression for the mathematical model presented in Additional file 2. The model estimates the RT-efficiency and the results are reasonable for five of the six genes. For Ins2 the efficiency is much lower than expected. The poor curve fit for the Ins2 gene results from the fact that the Ins2 data was generated for much higher copy numbers whereas our model for the PCR-noise was adapted for the low abundances of tge five other genes. (b) The proportion of the total noise for the PCR (filled symbols) and RT reactions (open symbols). Circles, squares and triangles are designated as in (a). The PCR and RT components do not have to add up to 1; the noise stemming from the dilution corresponds to the remaining noise. For Rps29, Chgb and Gfap, the PCR noise clearly dominates for all concentrations. For Ins2 and Sox2, the estimated RT efficiency is very low which means that this reaction will add a larger contribution to the total noise. Nes has an intermediate efficiency and for low copy numbers the RT noise dominates but it becomes smaller than the qPCR nosie when more transcripts are analyzed. Furthermore, the copy numbers are relatively high for Ins2 which deflates the PCR noise.

Mentions: Figure 3 shows that the total technical noise is very low at initial mRNA copy numbers down to ~100. The noise model allows us to estimate the RT-reaction efficiency (i.e. the fraction between the resulting number of cDNA copies and the initial number of mRNA copies) for each gene by non-linear regression. We find that the RT-reaction efficiency is 99%, 85%, 67%, 28%, 8% and 3% for Gfap, Rps29, ChgB, Nes, Sox2 and Ins2, respectively. The estimate for Ins2 is clearly much lower than what we would expect and we believe that this is due to insufficient data on the qPCR noise for high cDNA copy numbers for the fitting procedure (see Additional file 2). We conclude that for high RT-efficiencies, the noise originating from the RT reaction is comparatively low, 5–20% of total noise, while the qPCR noise dominates (30–90%) at low concentrations. If the RT-efficiency is low (<10%, i.e. Sox2 and Ins2), the qPCR noise is insignificant for all concentrations.


Quantification of mRNA in single cells and modelling of RT-qPCR induced noise.

Bengtsson M, Hemberg M, Rorsman P, Ståhlberg A - BMC Mol. Biol. (2008)

Noise levels for the measurements of known quantities of mRNA from Rps29, Chgb, Ins2, Gfap, Nes and Sox2. (a) The measured noise strength (η2 = SD2/mean2) for Rps29 (blue circles), Chgb (red squares) Ins2 (black triangles), Gfap (green triangles), Nes (cyan stars) and Sox2 (magenta diamonds) with corresponding fits (lines) obtained using non-linear regression for the mathematical model presented in Additional file 2. The model estimates the RT-efficiency and the results are reasonable for five of the six genes. For Ins2 the efficiency is much lower than expected. The poor curve fit for the Ins2 gene results from the fact that the Ins2 data was generated for much higher copy numbers whereas our model for the PCR-noise was adapted for the low abundances of tge five other genes. (b) The proportion of the total noise for the PCR (filled symbols) and RT reactions (open symbols). Circles, squares and triangles are designated as in (a). The PCR and RT components do not have to add up to 1; the noise stemming from the dilution corresponds to the remaining noise. For Rps29, Chgb and Gfap, the PCR noise clearly dominates for all concentrations. For Ins2 and Sox2, the estimated RT efficiency is very low which means that this reaction will add a larger contribution to the total noise. Nes has an intermediate efficiency and for low copy numbers the RT noise dominates but it becomes smaller than the qPCR nosie when more transcripts are analyzed. Furthermore, the copy numbers are relatively high for Ins2 which deflates the PCR noise.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2483285&req=5

Figure 3: Noise levels for the measurements of known quantities of mRNA from Rps29, Chgb, Ins2, Gfap, Nes and Sox2. (a) The measured noise strength (η2 = SD2/mean2) for Rps29 (blue circles), Chgb (red squares) Ins2 (black triangles), Gfap (green triangles), Nes (cyan stars) and Sox2 (magenta diamonds) with corresponding fits (lines) obtained using non-linear regression for the mathematical model presented in Additional file 2. The model estimates the RT-efficiency and the results are reasonable for five of the six genes. For Ins2 the efficiency is much lower than expected. The poor curve fit for the Ins2 gene results from the fact that the Ins2 data was generated for much higher copy numbers whereas our model for the PCR-noise was adapted for the low abundances of tge five other genes. (b) The proportion of the total noise for the PCR (filled symbols) and RT reactions (open symbols). Circles, squares and triangles are designated as in (a). The PCR and RT components do not have to add up to 1; the noise stemming from the dilution corresponds to the remaining noise. For Rps29, Chgb and Gfap, the PCR noise clearly dominates for all concentrations. For Ins2 and Sox2, the estimated RT efficiency is very low which means that this reaction will add a larger contribution to the total noise. Nes has an intermediate efficiency and for low copy numbers the RT noise dominates but it becomes smaller than the qPCR nosie when more transcripts are analyzed. Furthermore, the copy numbers are relatively high for Ins2 which deflates the PCR noise.
Mentions: Figure 3 shows that the total technical noise is very low at initial mRNA copy numbers down to ~100. The noise model allows us to estimate the RT-reaction efficiency (i.e. the fraction between the resulting number of cDNA copies and the initial number of mRNA copies) for each gene by non-linear regression. We find that the RT-reaction efficiency is 99%, 85%, 67%, 28%, 8% and 3% for Gfap, Rps29, ChgB, Nes, Sox2 and Ins2, respectively. The estimate for Ins2 is clearly much lower than what we would expect and we believe that this is due to insufficient data on the qPCR noise for high cDNA copy numbers for the fitting procedure (see Additional file 2). We conclude that for high RT-efficiencies, the noise originating from the RT reaction is comparatively low, 5–20% of total noise, while the qPCR noise dominates (30–90%) at low concentrations. If the RT-efficiency is low (<10%, i.e. Sox2 and Ins2), the qPCR noise is insignificant for all concentrations.

Bottom Line: The noise is insignificant for initial copy numbers >100 while at lower copy numbers the noise intrinsic of the PCR increases sharply, eventually obscuring quantitative measurements.Noise in single-cell RT-qPCR is insignificant compared to biological cell-to-cell variation in mRNA levels for medium and high abundance transcripts.To minimize the technical noise in single-cell RT-qPCR, the mRNA should be analyzed with a single RT reaction, and a single qPCR reaction per gene.

View Article: PubMed Central - HTML - PubMed

Affiliation: Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, The Churchill Hospital, Oxford, OX3 7LJ, UK. martin.bengtsson@med.lu.se

ABSTRACT

Background: Gene expression has a strong stochastic element resulting in highly variable mRNA levels between individual cells, even in a seemingly homogeneous cell population. Access to fundamental information about cellular mechanisms, such as correlated gene expression, motivates measurements of multiple genes in individual cells. Quantitative reverse transcription PCR (RT-qPCR) is the most accessible method which provides sufficiently accurate measurements of mRNA in single cells.

Results: Low concentration of guanidine thiocyanate was used to fully lyse single pancreatic beta-cells followed by RT-qPCR without the need for purification. The accuracy of the measurements was determined by a quantitative noise-model of the reverse transcription and PCR. The noise is insignificant for initial copy numbers >100 while at lower copy numbers the noise intrinsic of the PCR increases sharply, eventually obscuring quantitative measurements. Importantly, the model allows us to determine the RT efficiency without using artificial RNA as a standard. The experimental setup was applied on single endocrine cells, where the technical and biological noise levels were determined.

Conclusion: Noise in single-cell RT-qPCR is insignificant compared to biological cell-to-cell variation in mRNA levels for medium and high abundance transcripts. To minimize the technical noise in single-cell RT-qPCR, the mRNA should be analyzed with a single RT reaction, and a single qPCR reaction per gene.

Show MeSH
Related in: MedlinePlus