Skin swabbing for DNA sampling of zebrafish

WN: Our full protocol is available to read online. You can get all the information you need on setting up your equipment and the practical steps of swabbing within this published methods article. To begin with, we recommend practicing restraining and releasing a fish with a hand net, as per the published protocol. It can be a little tricky to hold the fish the first time, but it gets easier the more you do it. As with all new techniques, it takes some practice to perfect your method, but this initial time investment is worth it to minimise the impact that DNA sampling has on fish health and welfare.

GP: I would recommend starting with large, short-finned adults as this will make handling the fish a little easier and will allow you to gain confidence in carrying out the DNA collection technique.

When it comes to carrying out the procedure, make sure you have your all your equipment ready and ensure that you only bring the number of animals to your procedure space that you can comfortably process without compromising their welfare. For example, fish should not be kept in off-system holding tanks long enough for there to be changes in water temperature.

KFB: My recommendation would be to start with lines with simple genetic makeups that produce strong bands when using DNA extracted from fin clips. Lines that show clear separation of fragments in gel electrophoresis, have a high yield of DNA product, do not require restriction digest to distinguish individuals by genotype and have no more than a single mutation or transgene will make a good starting point. When establishing skin swabbing, I would also recommend doing fin clips and swabs from the same line (not necessarily the same fish) in parallel and comparing the results (e.g. PCR, gel electrophoresis, sequencing etc.). Where these results indicate weak performance for a certain line (i.e. they generally give weak or barely visible bands during electrophoresis and/or have a low PCR yield) our solution has been to set up the PCR using twice the volume we usually use, and this has worked well.

KFB: Yes, some protocols might need to be adjusted slightly when switching to DNA obtained from skin swabs.

In our facility we had optimized our PCR settings (from fin clips) so that we could use a really small reaction volume and when we switched to skin swabbing we found that initially we did not get the desired results. To rectify this, we found that doubling the reaction volume was sometimes all that was needed.

Whenever we have complicated PCRs that have a high failure rate due to factors such as low gene expression, difficult DNA structure and formation of hairpins etc. we sometimes have to adjust the PCR parameters. In general, our PCR settings are chosen to give us the best possible result, starting with a line-specific DNA concentration. If starting with a lower concentration, the PCR settings might have to be optimized for good results again and just using the standard protocol might not be sufficient.

I would emphasize that one cannot assume that whatever the standard genotyping protocol for a given line is/was using fin clips, this can be copied one-to-one using skin swabbing. It might be necessary to make adjustments to your protocol when using DNA extracted from skin-swab samples, as we have done in our facility. However, in terms of refinement, this is absolutely worth the effort.

RK: Initially we struggled with high variability from swab samples but we have resolved this by making changes to how rapidly we prepare and process them. Since implementing this approach all the samples have given good DNA allowing for successful genotyping of all the genes tested so far.

Our approach is to take five swab samples, pause to begin the extraction process and then continue to collect another around of swabs. With two people we find that this method efficient, enabling seamless collection of samples.

WN: We restrain the fish using a combination of a hand net (e.g. a Marina Fine Soft Mesh Fish Net or similar) and wetted sponge with a groove cut into it (e.g. a Vitrex 10 2904 square sponge or similar). The thumb and forefinger can be used to hold the fish still, with the net covering the head and the tail of the animal (see the below image for an example of this). The fish’s body is placed into the groove on the sponge, which helps to keep it still. Some people prefer to just use a flat wetted sponge, with no groove [9], or single-use wet gauze rather than a sponge [10] – the technique that you find quickest and most practical will be down to your personal preference.

The aim is to take your sample and get the fish back into water as quickly as possible, so make sure that everything is ready to go before you net your fish out of the water. If you are not used to restraining small fish, we recommend practicing safe and efficient restraint before you have a go at swabbing. You’ll gain confidence quickly and will find what works for you.

GP: A common concern I hear is that handling smaller-bodied fish is challenging without sedation. If this is something you are struggling with initially, it may be advisable to start with slightly larger animals before progressing onto handling smaller fish once your confidence with the technique has grown.

To help the procedure run smoothly and to keep the fish calm, use a quiet space and avoid noisy rooms, vibrations and areas with high footfall. Fish can find exposure to bright lights stressful, which can influence how jumpy they are. Covering the holding container is often an effective way to reduce stress before you begin a procedure.



If you have practiced the technique and are taking reasonable steps to keep the fish calm handling unanaesthetised fish should become easier the more you do it. If you continue to struggle with manual restraint, for example, when handling a particularly hyperactive line, you may conclude that swabbing without sedation is not the best option for you. It is important to be open minded about what is possible with new techniques, but it is also essential to be pragmatic about what will work for your own situation. The onus is on us as researchers to refine our methods where possible; trialling skin swabbing will allow you to see whether it is a refinement appropriate for your individual circumstances.

RK: We simply use small nets and make sure the head is shaded to encourage calmness - I think covering eyes really makes a difference.

WN: Very little pressure is required when swabbing the fish, to prevent any possible discomfort or damage to the animal. The mucus can be collected readily by gently stroking the fish’s flank starting below the gills and moving towards the tail (this direction is important for fish with scales, such as zebrafish).

GP: Only light pressure is required, and the swab should be moved along the fish’s flank towards the tail in one clean movement. Avoid swabbing the operculum/gill area. I would suggest starting one swab width posterior of the gill plate, and even starting half way down the body will suffice if this means avoiding unnecessary stress by having to reposition the fish. On reaching the tail, lift the swab clear of the fish and repeat to retrieve enough material. For long-finned fish, you may need to gently clear the pectoral fin away from the body before collecting your DNA sample. This can be done by using the same swab.

WN: We have reliably collected DNA from fish ≥ 20 mm using the skin swabbing technique [9]. We are exploring whether swabbing zebrafish as small as ≥ 10 mm is viable, however this requires careful investigation and so we do not currently recommend the technique for fish smaller than 20 mm.

WN: We have successfully used both commercially available rayon-tipped swabs, and simple cotton buds bought in a pharmacy or supermarket. While these are consumable items, they are easy to obtain for a very low cost. The swabs do not need to be individually wrapped, but they should be autoclaved and dried before use. We are hoping to trial the use of smaller-tipped swabs that can be used in 96-well plates.

WN: It does not matter if the swab gets a little wet when collecting mucus from the fish’s flank. This will not affect subsequent DNA extraction.

WN: Skin swabbing and fin clipping appear to be equally useful when amplifying genes to identify fish. DNA collection by skin swabbing tends to lead to a lower of concentration of DNA being extracted compared to fin clipping. However, our research has shown that the purity of the DNA is similar in both techniques [9].

We routinely use skin swabbing to identify zebrafish strains by PCR in our laboratory. We found that once we had familiarised ourselves with the skin swabbing sampling technique, we had very few samples that we could not amplify by PCR or sequence. We see a similar level of success when sequencing PCR reactions that used fin clip DNA as a template, but it may be that some practice is required to optimise the skin swabbing technique in your facility.

DNA from skin swabbing has been used to successfully identify fish by PCR in peer-reviewed literature [9-14]. The zebrafish strains included in these studies were AB wild type, casper, Tg(vmat2:GFP), Tg(hb9:GFP), Tg(glyt2:GFP) [9]; NHGRI-1, kca33Tg, and kca66Tg [10]. Sex markers have been also been used to differentiate between male and female stickleback [9], medaka [13] and wild roach [14] using DNA collected by skin swabbing.

We have not tried to use DNA extracted from skin swab samples for RAD-TAG sequencing or whole genome sequencing, so we cannot comment on the suitability of swabbing for these purposes at the present time. It is something that we would like to investigate in the future.

GP: You might find that swabbing is not suitable for all purposes, but if it works in 90% of cases that is already an improvement in terms of reducing the number of animals that go through this invasive procedure worldwide. 

WN: Skin swabbing collects skin cells that have been shed by the fish and are trapped in the fish’s mucus layer.

WN: Studies show that cross contamination of this kind is not an issue [9,12]. We have demonstrated that cross contamination does not occur between zebrafish, even when the fish have been living at high densities and have been in direct contact with each other shortly before being swabbed [9].

WN: No, skin swabbing has been successfully used as less invasive method for DNA sampling of fishes with a body length ≥ 30 mm [12,13], amphibians [15] and mammals [16]. Our research has validated skin swabbing for DNA sampling in smaller-bodied (≥ 20 mm) laboratory fishes. Typically fin clipping is used as standard to obtain DNA samples from laboratory fishes and swabbing presents an opportunity for a refinement.

GP: My colleague at the University of Exeter, Dr Anke Lange, has successfully used skin swabbing for determining the genetic sex of roach (juvenile and adults) and stickleback. Within the same research group, they have also successfully used skin swabbing for metabarcoding the skin microbiome of tilapia [17].

WN: In our hands, we found that DNA extraction using standard molecular biology methods (phenol:chloroform extraction) worked better than commercial kits for skin swab mucus samples obtained from zebrafish and stickleback [9]. Other researchers have combined the skin swabbing protocol with a HotShot DNA extraction [10]. 

WN: Our research looked at behavioural and physiological indicators of stress and found that, overall, swabbing has a less profound effect on fish than collecting DNA by fin clipping [8]. For example, we found that removal of tissue during fin clipping led to a significant increase in the release of cortisol, the stress hormone, whereas when fish were swabbed the waterborne cortisol levels were not significantly different from control undisturbed fish.

Part of the negative welfare effect of fin clipping comes from the necessary requirement to anaesthetise the animals for the procedure. For example, MS222 (tricaine) treatment altered the opercular beat rate, a measure of ventilation in both zebrafish and sticklebacks [8] and commonly used anaesthetic agents have been shown to be aversive to zebrafish [17,18].

Swabbing is a less invasive procedure and does not require fish to be anaesthetised. The process of being netted and taken into the air for swabbing is stressful to the fish, but by reducing the stress associated with the sampling procedure (i.e. swabbing rather than fin clipping) the cumulative suffering of the animals is reduced.

WN: Any procedure that involves netting a fish, bringing it into the air and handling it is going to cause the animal some distress. It is important that we are always looking for ways to refine procedures to minimise the overall suffering that the fish experience. From our research findings, we are confident that using skin swabbing instead of fin clipping will reduce the cumulative suffering that laboratory fish experience.

We are investigating ways to further refine swabbing, starting with understanding more about how much mucus is removed and how quickly the mucus layer recovers. We have not observed any secondary infections or evidence of damage to the fish we have swabbed, but mucus is important for fish to protect against bacteria and pathogens so we want to learn more about this. As a cautionary measure we looked at pre-treating the fish with an analgesic dose of lidocaine (2mg/l), but preliminary data on this indicates that analgesia is not necessary for swabbing to be a refinement to fin clipping.

WN: Within this published protocol we used an analgesic dose of lidocaine (2mg/l) but did not place the fish under anaesthesia. Lidocaine is effective for reducing pain in zebrafish and is used in some fin clipping protocols, so we were motivated to explore pre-treatment with an analgesic as potential further refinement to the swabbing procedure. This was purely a cautionary approach while we investigated whether there was any evidence that the fish experience any discomfort from swabbing. Upon further investigation it does not appear that analgesia is necessary for swabbing to be a refinement.

WN: Our research shows that the data collected from swabbed fish, which measured cortisol release, behavioural indicators of anxiety and gene expression, was less variable than in fin clipped fish. This creates an additional 3Rs benefit as fewer animals are required for sufficient statistical power. From a practical perspective, skin swabbing alone is not a licenced procedure (in the UK) and our laboratory personnel prefer using swabs to scalpels as there is less risk of harm to the user.

GP: There are also environmental benefits of not using anaesthesia; when anaesthetic agents are disposed of down the sink, they have the potential to affect ecological systems.

WN: We recognise that experienced personnel are able to carry out fin clipping efficiently and that facilities are committed to minimising harm to their animals, but we are only able to comment on what the data shows us. There is strong evidence that commonly used anaesthetics are aversive to fish [18-22] and that the fin clipping procedure itself causes stress and pain [1-7]. Skin swabbing is still a procedure that requires care; however, our data show a reduced impact on fish welfare [8] and we believe that this presents an opportunity to refine DNA collection in laboratory fishes.

Our end goal is improve the welfare of laboratory fishes and we invite others to share any data they have to highlight opportunities for refinement, either directly with myself (whjn1@le.ac.uk) or with the NC3Rs (tech3rs@nc3rs.org.uk).

WN: If you would like help with a training programme, we’d be happy to provide support (whjn1@le.ac.uk). In the meantime, use the most refined anaesthetic protocol in terms of safety and aversive properties [18-22], ensure that analgesia is used to manage pain before and after the procedure [2,20,21] and allow fish to recover in visual contact with other fishes [23] (e.g. by pushing tanks together so fish can see each other or by partitioning tanks so that fish are separated but have visual and olfactory contact).

GP: Also consider how much tissue needs to be removed from the fish for your purpose and do not take more of the fin than is needed. I have spoken to zebrafish users who have reviewed their fin clipping protocols and found that only a very tiny piece of tissue (e.g. 1-2mm² equivalent to the tip of one tail lobe) was needed to complete >10 PCR runs. Fin clipping tail lobes back to the scissor of the tail was found to be completely unnecessary. The smaller the piece of tissue taken, the faster the recovery of the fish and sooner it can be returned to its home tank.