Detection of circulating breast cancer cells by reverse transcriptase polymerase chain reaction (RT-PCR
Kalpit Shah
SID: 18930499
Summary:
The relationship between circulating tumor cells and the development of secondary disease in a breast cancer patient is not fully understood. This paper talks about a method to detect small numbers of such cells in peripheral blood (PB), hoping that it may aid treatment and diagnosis. The authors have developed a method that uses RT-PCR. The benefits include the necessity of a very small amount of blood and also more sensitivity than the other current methods.
PCR has traditionally used to detect DNA abnormalities commonly seen in cancer cells. However, DNA amplification doesn’t discriminate between viable and non-viable carcinoma cells. Hence RT-PCR is ideal as it detects and amplifies the presence of specific mRNA’s only seen in malignant breast cancer cells. Here is a list of markers that have been seen in literature that are specific for viable breast cancer cells.
CK19 – problems of false positively seen when nested RT-PCR is used also if amplification of greater than 40 cycles is used because of the presence of pseudogenes. Hence it is recommended to amplify for less than 40 cycles and use primers that differentiate between the pseudo and the real gene.
CK20 – is associated with patients with an overall lower survival rate, but not always detected in all samples. Hence it isn’t reliable.
β-hcg – only seen in 12-18% of breast carcinoma. Also, because of the presence of different forms of it being present, it’s role is not well known. Additionally, though it is detected 80% of the time in the tissue of patients, it isn’t reliability detected in the peripheral blood.
Problems of RT PCR:
1) False Positive tests – Because of the sensitivity of RT-PCR, illegitimate copies of unrelated and insignificant genes in the test cells lead to a positive solution. Researchers have gone around this problem by using 100 cells or less for RT-PCR and also using more specific primer sequences. It is recommended to use of primers that span at least one intron. This can also act as control since you’ll get 2 bands – one from the RNA and one from genomic DNA is contamination occurs. To avoid the problem of psuedogenes, it is recommended to use primers that include differences in the two genes.
2) False Negative – Certain Reagents are better are RNA extraction – Trizol LS is shown to be the best. Because carcinoma cells are shed into the blood stream intermittently, multiple sampling of blood at various times is recommended.
3) Multiple markers is shown to lead to a more specific match of the genes in question and also protects against the problems mentioned above.
10 comments:
"Hence RT-PCR is ideal as it detects and amplifies the presence of specific mRNA’s only seen in malignant breast cancer cells."
I can believe that specific mRNAs are overexpressed by breast cancer cells, but are they only seen there? These genes must have a purpose in some kind of healthy cell.
You said "It is recommended to use of primers that span at least one intron. This can also act as control since you’ll get 2 bands – one from the RNA and one from genomic DNA is contamination occurs."
Why would you get a band from the gDNA? If one of your primers spans and exon-exon boundary, it should not be able to amplify anything right? Isn't that the point of designing primers that span an exon-exon boundary?
Why don't they choose to use qRT-PCR because RNA is not gonna last very long, depending on the time points in which RNA is extracted, one might get unlucky and get less RNA than a normal breast cancer cell should yield...
Their conclusion was that RT-PCR now can only test for a true negative using CK 19 and B-hcg. But you said it is only about 80 percent of cancer patient express B-hcg and this gene can not be trusted if collected from periphery blood instead of from tumorous tissue. Then their conclusion that patient is cancer free in the absence of B-hcg in periphery blood is really not valid. Then I guess even though the RT-PCR has higher sensitivity, no current method have the specificity it requires to output some reliable results.
This is a very interesting way of looking for diagnostic tools. It would be very useful even if the genes exist in other cells, but are way upregulated in malignant breast cancer cells. It's clearly a new idea, but I think it's a very interesting direction to go in.
As you stated in the analysis, because of the extremely sensitive nature of RT-PCR, illegitimate copies of insignificant genes may show up. You say that in order to solve this, the researches selected 100 cells or less. Did the author show a comparison of using other number of cells (like a comparison with using 50 cells vs 500 cells) and analyzing these results?
Terry: You're right. As we found out through our project, HeLa cells (used in literature as control) did express HER2, which was supposed to be a breast cancer biomarker. qRT PCR would have been a better choice.
Tim: You're right. Intron-spanning primers would get rid of any amplification of the gDNA. What I was actually meant was if one of the primers was on exon1 and the other one was on exon2 then you'd get two bands. One from the processed mRNA (smaller) and one from the genomic DNA (larger) in the case of contamination. Am I correct in assuming that?
Andrew: qRTPCR is better because we lets you get an idea of the initial concentration of RNA. I'm not sure if RNA degradation would be a problem as qRT PCR also uses RNA.
Sky: I agree, but 80% sure is better than no result at all.
Neeraj: You raise an excellent point. From my understanding they didn't. My guess would be that somewhere in literature its been shown that about a 100 cells are ideal to detect specific genes.
Yeah, Kal, I think you're right. By having your primer span at least the length of an intron (though not too big), you separate your desired cDNA from any other DNA. That way, if there's contamination in the lane, you'll get two separate bands and can isolate them.
You stated the list of markers that were specific for viable breast cancer cells. Is the paper certain that some of those genes aren't expressed in other cell types, especially other types of cancer cells? Did the paper go about explaining how they isolated them, other than simply referencing towards literature?
Otherwise -- what a novel way to explore diagnosis methods using already well known lab techniques, though I would have to agree with Andrew and Terry that qRT-PCR would probably been a more viable choice. I wonder if they did any follow up papers exploring that.
Yeah, I'm not sure RT-PCR would be ideal because the genes might be expressed in lower quantities, just not as much as in cancer cells. It would probably amplify it in both cases, so you wouldn't get a clear delineation between cancer and normal cells, unless perhaps you did a baseline to ensure equal concentrations of RNA or something along those lines.
Did they talk about low expression of bhcg in other cell types, can could they be amplified to give a false positive?
Post a Comment