Lab skills quiz – RNA/DNA techniques

 As the amount of protein in a sample increases, the absorption at 280nm increases therefore A260/A280 decreases. The more protein, the lower the ratio will be

Organic substances absorb best at 230nm therefore you would use an A260:230 to assess this.

To be fair, you would learn this while taking A260/280 measurements; the absorbance at 260nm on its own tells you about quantity of substance presence. However, the A260/280 ratio provides you additional information that you also should be recording whenever you measure your nucleic acids.

A280 can be used to measure protein concs… this is a clue for what the correct answer is!

This is quite low, it suggests your sample is contaminated with protein. Those proteins will inhibit your PCR reaction and give you false readings.

Pure RNA is generally accepted as having an A260/280 ratio of 2.0. You have to decide how pure is good enough for your experiments. 1.8 is quite commonly used but you might need higher or you might need to justify lower (e.g. very rare samples). Whatever you decide as your cut off needs to be made clear in your methods section as it is relevant to the interpretation of your results.

 

1 μg = 1000 ng, if you 1000/1500 = 2/3 = 0.67

This would give a 1% gel which is about right for this size of product. Realistically 0.75% to 2% would all be OK. You should choose your gel percentages so that they allow good separation around the point where you are interested

A 0.2% gel would be very weak and likely rip apart in your hands

1g in 10ml = 10 %, 0.5 ml in 50 ml = 1/100 so 0.1% final concentration. It won’t form a gel at all!

Got it. TRIZOL can be used to collect RNA, DNA and protein. If you guessed this, click on the other two to see what’s where!

The fluffiness you see here is the DNA

If you do this you will get bother the DNA and the RNA

This contains the proteins and other stuff like sugars.

  Ethidium Bromide; binds to DNA and gives a fluorescence signal when exposed to UV light. There are a number of dyes that work in similar ways that are supposed to be a bit safer, things like Midori green, Sybr safe etc. Whether they actually are safer is up for debate so take care!

this is a protein binding dye, you would use for total protein stains on a SDS-PAGE gel.

this quite a generic blue dye, often it used as a cell stain in things like adhesion assays.

this is a protein dye. Usually it is used to stain proteins on a membrane in western blotting type applications, for example to demonstrate equal loading.

may be elsewhere in methods which would be ok, but needs to be clear somewhere

maybe in the table but information about what you are/aren’t amplifying is important so should be somewhere!

 

RT-PCR allow you to amplify specific regions of DNA. You use primers and amplify the region in between, you don’t necessarily amplify the whole transcript. In a Northern blot, your probe binds to the RNA species at the size where it has run on a gel. If you want to know where a transcript starts or finishes you could consider a technique called RACE – rapid amplification of cDNA ends – this is a PCR-like technique but, again, will only let you know the length of transcript downstream/upstream of the priming point (and assumes that you have amplified through secondary structure). Select all that apply

A Northern blot can do this, but quantitative PCR will do this too (likely with less variation between repeats). Northern blots were what was used before qRT-PCR so you will see northern data in the scientific literature.

You can design primers that are specific to different isoforms and use them for qRT-PCR to generate relative abundance; however, most conventional qRT-PCR applications are limited to amplicons of about 200bp so you might struggle to generate efficient primers for longer transcripts. A Norther could help you here. There are also alternate techniques that involve probe binding in similar ways; you may be able to design probes that bind across specific exon junctions, however some highly alternatively spliced genes may have the same exon-exon junction in multiple transcripts, a Northern might allow you to look at all the options in one go.

You can do this by Northern blotting, but using a mix of different reverse transcription primers could also allow you to do this by RT-PCR.

Northern blots are a way of detecting transcripts, try again

Chi squared tests are for comparing categorical variables.

t tests are designed for comparing two categories only, here you have three treatment groups

This looks like the correct answer as you have 3 treatment groups and continuous outcome variables BUT this approach doesn’t account for the plan to have 5 different outcome variables; the 5 transcripts. If you don’t plan for the multiple comparisons then your study runs the risk of being underpowered.

You are planning multiple comparisons across four groups therefore your sample size calculation should plan for multiple comparisons. a Multivariate ANOVA(MANOVA) might be your best bet. Read about MANOVA here.

assuming your experimental techniques are well optimised this variability will always be the biggest variability in human studies. Experimental “n” will be decided here.

Ideally the sample will be immediately immersed in liq nitrogen upon removal and storage times will be short. However, even with very careful treatment, there is quite a large scope for differences here; RNA is very sensitive and will degrade easily, small differences here will carry right through your set up

there are commercial kits available that make this step more consistent but when you are working with chunks of tissue there will have to be some homogenisation steps which gives you a large source of variability… RNA is sensitive to lots of things, shorter/longer or incomplete digestion would be a problem. Any difference in RNA isolation will carry all the way through the rest of your experiments, e.g. inhibitors of PCR will be differentially present.

This step is usually quite consistent compared to the others, there will be variability introduced here but it likely won’t be as big as some of the others. You can determine how much by doing two separate RT reactions with the same source RNA and comparing the results thereafter

There will be variability here but it will arise from in-precise pipetting rather than biological variability. The same cDNA should give you very similar results

Although there will be differences here but they should be small. You can design in controls into each run to confirm that there has not been a problem

Yup, the simple answer is definitely possible! Unfortunately, it’s not the only explanation.

Yes, you don’t know yet if your samples are any good. A simple explanation is that they aren’t! Run another PCR with for a gene you know will be present to check DNA quality.

The PCR worked for the positive control, therefore you don’t have reason to believer the primers are “off”

Linear DNAs (such as PCR products) run at their size, so if you have a well resolved gel you can look to the size for interpretation purposes. Most genes are alternatively spliced in some way, here the exons could be organised in a variety of ways with different sized products (see below). Although this interpretation makes sense I would still check e.g. by sequencing the extra bands.

While making the cDNA you use very little RNA and also include an RNAse step to digest it, it’s very unlikely that these RNA species would be in your final reaction and, even if they were, they wouldn’t be very abundant and wouldn’t run at these sizes.

For this to work as an explanation, your PCR primers would have to recognise a coupe of places in the rabbit sequence. Technically possible but not the most likely explanation of these data

Although you could have genomic DNA contamination, a PCR using these primers would give a very long product rather than the shorter products observed here

The two primer sequences are still present in the cDNA within your sample (otherwise you would get nothing).  If all your RNA had degraded you would lose all the bands in this experiment. Partial degradation would unlikely to have cut out the parts between the two primers. Good guess though! Smaller molecular weight bands in western blots can sometimes be explanation by partial digestion.

Endpoint RT-PCR won’t give you this sort of quantitative data

Yes, this is a plausible answer, the labelling isn’t brilliant (see Q2) but this is the sort of graph that could be generated from those type of data

The “fold change” part means that there is some quantitative measurements being compared. Plausibly it could be a measurement of mutational frequency i.e. #mutations/kbp and plotted as fold change in mutational rate.

  Yes, this graph could represent those data. Better labelling could have removed ambiguity, see Q2.

This helps the reader interpret the data, it’s part of the results. The details could be in the figure legend but adding a few words to the y axis would make the information clearer.

Not a big ask! I don’t want to have to hunt for this information, make it clear and easy to see

Hunting for details in the figure legend or, worse, elsewhere is annoying! Labelling the treatments makes the data easier to understand

You could ask for this information to be included in addition to the fold change if you felt it was relevant to the story. However, if the question this experiment was designed to address related to change in expression relative to untreated then an absolute abundance graph could be harder to interpret, especially with the relatively small changes.

May not necessarily be required, but if, for example, Figure 1B referred to the same data but at the protein level then this addition could help remove ambiguity.

I agree, there probably is no advantage of colour here. However, it’s probably OK. If the colour has potential to influence the data interpretation is should be changed. Similarly, if the colour choice wasn’t great for colour blind people then it should be changed. Here the colour doesn’t add anything or influence the interpretation. (In case you are wondering I can’t work out how to change the colours from green and red to signify right and wrong in this quiz!)

You could request this change if you want, but a log scale is appropriate for these data so I am not sure what the justification would be. In the graphs below I have made the blue treatment double the expression level, and the red treatment halve the expression level. The scale of the change is the same but in different directions. You can see how the same data would look on different types of graph. When making your figure, you have to decide what tells the story of your data in the most accessible and easiest to see way.

DMSO should have very little effect, fluastin should increase expression dramatically, C notibimin should decrease expression and zoombastin could have any effect at all. B and D could go the other way round

Yes, you aren’t measuring mRNA degradation but rather relative mRNA abundance. Doing the experiment in this way, you would infer from a change in level that the difference was due to differential degradation but it could instead be that the treatments have affected the expression of other things, the mRNA changes could be due to changes in transcriptional rates.

Nope, This experiment would examine PAX2 mRNA abundance not degradation

Quick, cheap and easy! All you really need is a few sets of primers and some miRNA mimics. You can buy these, test the primers and then do the experiment. So although this experiment isn’t perfect it might be the best place to start. Note: I would adjust my hypothesis here to more accurately reflect what my experiment was testing.

A few options here. You could inhibit transcription (using actinomyosin D) and do a time course experiment isolating RNAs at different points after both treatments. Then you would compare the rate at which PAX2 levels dropped. Again this is still indirect, other factors could be at play, but it would be a bit better as it would remove one alternative interpretation. Another option would be to generate a reporter construct for example using a coloured protein or luciferase gene followed by the 3′ UTR of PAX2. This would take some time and effort. Ideally you would then mutate the putative miRNA binding sites (seed regions) and then compare the effects of the mimic treatments on the decay rates. This experiment would tell you if your miRNA was capable of degrading the mRNA but it doesn’t mean that it does so in normal conditions. There are some other options but these three approaches are quite commonly used in the literature. Likely the  answer is to use a combination of approaches and get the best answer you can. It’s rare that you will find the perfect way to do any experiment so this sort of triangulation is usually what you need to do.