UCL Division of Biosciences



Stern, C.D. (1999) Grafting of somites. In: Molecular Embryology: Methods and Protocols. (P.T. Sharpe and I. Mason, eds.). New Jersey: Humana Press. pp. 255-264. (Meth. Mol. Biol. 97, 255-264).

Grafting of somites

Claudio D. Stern

Department of Genetics and Development

College of Physicians & Surgeons of Columbia University

701 West 168th Street #1602

New York, NY 10032, U.S.A.


The somites are an intriguing invention of vertebrate embryos. They represent the most overtly segmented structures of the body plan, but they give rise to both obviously segmental (e.g. the axial skeleton) as well as not-so-obviously metameric (dermis and skeletal muscle) elements. In addition, they play a key role in controlling several aspects of the organization of the central and peripheral nervous system of the trunk, and appear to participate in several different types of inductive interactions both within themselves and with neighbouring tissues like the neural tube, the notochord, the metanephric and lateral plate mesoderm and the ectoderm and endoderm (see 1, 2 for reviews).

Questions that can be addressed by manipulating somites range from investigations on the mechanisms by which metameric pattern is established, to their influence on the segmental outgrowth and differentiation of precursors of the peripheral nervous system (neural crest cells, motor axon growth cones), to the control of myogenesis, and patterning and the establishment of regional identities of cells that contribute to the dermis, limbs and axial skeleton. Previous experiments (1, 2) have suggested that although many aspects of somite development are controlled by surrounding tissues, many others appear to be remarkably autonomous.

Somites form at the posterior end of the embryo, such that a pair of somites is added every 100 minutes or so (1, 2 for reviews). Therefore at any particular stage of development, the embryo contains younger (more recently formed) somites at their caudal end, and older (at more advanced stages following their formation) more rostrally. To indicate this, Ordahl and his colleagues have introduced a "somite stage" numbering system, using Roman numerals to indicate the position of the somite being referred to with respect to its neighbours (see 2). In this system, somites are counted upwards from the most recently formed one, which is designated as I. The most caudal 4-6 somites are usually epithelial spheres. Stages V/VI and higher designate somites whose dorsolateral surfaces still remain epithelial (the dermomyotome; Fig. 1) but whose ventromedial parts have become mesenchymal once more, to form the sclerotome (1, 2). The neural tube, notochord, ectoderm and endoderm all play a role in determining the dorsoventral polarity of the somites with respect to their ability to form a dermomyotome and a sclerotome.

Despite the simplicity of this numbering system, it is important to remember when investigating somite development that overlapped with this age-structure is also a position-dependent address (reviewed in 1, 2). This can be demonstrated by transplanting somitic mesoderm from the thoracic level to the neck, where they go on to develop ribs as if they had not been transplanted. However, if a similar experiment is conducted to investigate the nature of the muscles that develop, it is found that any somite will give rise to muscles appropriate for their new position. Thus, some somitic derivatives behave in a cell autonomous way concerning their positional information, while others are subject to cues from their environment. In addition, the most rostral 5 or so somites ("occipital somites") have a different fate from the rest (they do not contribute to the vertebral column and some of their cells appear to contribute to the tongue) and in addition do not support the development of dorsal root ganglia from neural crest cells migrating in them.

Experiments in which the somitic mesoderm is manipulated can be done either in ovo or in whole embryo (3-5) culture. The main advantage of the former method is that it allows embryos to develop for a long time, even up to hatching, but embryos younger than about the 4 somite stage (Hamburger & Hamilton, (6); [HH] stage 8) are very difficult to manipulate in this way and their survival is poor. The latter technique allows very young embryos (even before incubation, at pre-primitive streak stages) to be operated, but they will only survive for 36-48 hours, even in the most expert of hands. In the following sections, I will describe two examples. The first is a detailed method for operations on somitic mesoderm in ovo, in which the anterior half of the segmental plate (unsegmented paraxial mesoderm; see Fig. 1) of a quail embryo is grafted into the same position of a host chick embryo. The same procedure is "generic" and can be adapted easily for manipulation of newly-formed or older somites at stages 9-15, as well as for manipulations of the notochord, neural tube and other tissues at these stages. The second example, to be used in conjunction with the instructions in the chapter on grafting Hensen's node (5), gives advice on manipulating younger embryos to investigate the mechanism of segmentation from the primitive streak stage onwards (see 7, 8). The procedures given here have been adapted from those in (9).


2.1 Materials

1. dissecting kit: 2 pairs small forceps (watchmakers'; number 4 or 5), 1 pair small scissors (about 2 cm straight blades), 1 scalpel (No. 3 handle, No. 11 blade), a Gilson micropipette for 3 µl, yellow tip(s)

2. eye surgeon's micro-knife, 15E angle. Suitable microknives are: Micro-feather microsurgery scalpels for eye surgery, 15E blade angle, Catalog number 715, manufactured by Feather (Japan) and marketed by pfm GmbH. Sold in boxes of 5, not cheap!

3. 2 entomological pins, size A1 or D1, or sharpened tungsten needles, each mounted by melting the end of a short Pasteur pipette (to act as handle) or inserted into a metal needle holder.

4. plasticine (modelling clay) to make ring for resting egg on its side

5. 100 ml Calcium-/Magnesium-free Tyrode's saline (CMF):

dissolve: 80 g NaCl, 2 g KCl, 0.5 g NaH2PO4.2H2O and 10 g glucose in 1 litre H2O. Autoclave for storage.

On day of use, dilute 1:10 with distilled water. The working solution may be buffered with bicarbonate, but we usually omit this.

6. antibiotic/antimycotic solution, 100x concentrate (e.g. Sigma A9909)

7. 70% ethanol to wipe shell

8. PVC tape to seal egg

9. 2 Pasteur pipettes and rubber teats

10. container for egg waste

11. 10 ml plastic syringe with Gilson yellow tip stuck into the end, filled with high vacuum silicon grease or vaseline.

12. 1 ml syringe, 27G (or finer) x 3/4" needle (for ink injection)

13. 1 ml syringe, 21G needle (for antibiotics)

14. 5 ml syringe, 21G needle (for withdrawing albumen)

15. paper tissues

16. 35 mm plastic dish coated with Sylgard and steel insect pins, size A1 or D1

Sylgard 184 (Dow Corning) is clear silicone rubber polymerized by mixing two components (9 parts rubber solution : 1 part of accelerator/catalyst). Mix the two well and pour to the desired depth (2-5 mm) into the plastic Petri dish. Allow the dishes to stand for about 1 h at room temperature for air bubbles to leave, then cure at about 55EC until polymerised (3 h to overnight). The dishes can be stored indefinitely. Black Sylgard is also available.

17. Indian ink (Pelikan Fount India is best; most other makes are toxic), diluted 1:10 with CMF and loaded in the ink injection syringe

18. 50 ml Trypsin (DIFCO, 1:250), freshly made up to 0.12% (w/v) in CMF

19. hens' and quails' eggs incubated 40-44 h so that they are at stage 10-12. The hens' eggs (hosts) should have been resting on their sides at least 30 minutes.

20. dissecting microscope, preferrably with transmitted light base

21. fibre optic incident illumination

2.2. Method

Here the host embryo is operated in ovo, which is only suitable with ease for embryos older than about 36-48 h incubation. As an example, the operation described here consists of a graft of presomitic mesoderm (segmental plate) with a quail embryo as the donor and a chick embryo as the host. An identical can be followed to transplant paraxial mesoderm that has already formed somites, but it should be remembered that this is much easier when it involves the youngest (most caudal) 5-6 pairs of somites at stages 10-14. Somites situated more anteriorly in the embryo have already separated into dermomyotome and sclerotome (Fig. 1) and special care is needed both to keep these components together during grafting and to separate them cleanly. Whatever the operation to be performed, it is generally a good idea to fix for histological analysis some of the pieces like those to be transplanted, as well as some host embryos immediately after the operation, to confirm that the graft involves the tissues of interest and no contaminating cells.

2.2.1. Preparation of donor quail embryo

1. Remove quail eggs from incubator. With the scissors, gently tap near the blunt end of an egg so as to penetrate the shell. Use the tip of the scissors to cut off a small cap of shell near this end, carefully so as to avoid damaging the yolk.

2. Allow egg white to pour into waste bucket, assisted by the scissors, taking care to avoid damage to the yolk. You may need occasionally to cut through the rather thick albumen using the scissors.

3. Once most of the albumen has been poured off, make sure the embryo is uppermost; if not, turn the yolk by stroking it very gently with the sides of the scissors.

4. Use the scissors to make 4 cuts into the vitelline membrane around the embryo. If the embryo does not lie exactly in the centre of the egg, make the first cut on the side of the embryo nearest the shell, and proceed in this way until all 4 cuts have been made. Make sure all the cuts meet each other.

5. Pick up the square of embryo/membrane by grasping one corner with fine forceps and transfer it immediately to the Sylgard-coated 35 mm dish containing about 5 ml CMF-Tyrode's solution.

6. Using two pairs of fine forceps, separate the vitelline membrane from the embryo, but leave the extraembryonic membranes attached to the embryo.

6. Stretch out the embryo (either side up) by placing 4-6 entomological pins (size A1 or D1) through the corners of extraembryonic membranes, into the Sylgard rubber. The embryo should be under some tension. Put the dish aside while preparing the host.

2.2.2. Preparation of chick host embryo

The procedure described here differs from that in the chapters on grafting of the notochord, neural tube, AER/ZPA and neural crest (10-13). In the method described here, a small window is cut into the shell and the embryo floated up to the level of this window for the operation. The advantages of this technique are many: (a) the embryo lies very close to the operator's hands, (b) it is completely submerged in liquid at all times (which avoids drying out and simplifies some of the manipulations), (c) the liquid above the embryo can be changed several times, for example to wash the embryo, (d) that the light illuminating it can be shone tangentially to its surface (the bubble of liquid above acts as a lens that greatly enhances the optical clarity), (e) the tension generated by floating the embryo seems to aid the action of the trypsin, such that tissues almost appear to dissect themselves, (f) it is very easy to notice if the endoderm has been punctured accidentally because ink will fountain out very quickly and (g) that after the operation the small hole can be closed very tightly with plastic (PVC, or electrical) tape, which allows the egg to be incubated with the window downwards and turned - both of these greatly enhance survival.

1. Shape plasticine (modelling clay) into a ring about 2" (5 cm) in diameter and place it on the stage of the microscope. Place a hens' egg (host) onto the plasticine ring, being careful that it is not rotated it with respect to its resting position.

2. Using the 5 ml syringe with 21G needle, held nearly vertical, insert needle into blunt end of egg until the shell is felt at the bottom surface. Withdraw 0.5-1 ml egg albumen, which should come up easily.

3. Score a shallow 1 cm x 1 cm square on the top of the shell with the scalpel and lift up the square of shell.

5. With a pair of watchmakers' forceps, pierce and remove the underlying shell membrane, after wetting it with CMF. Avoid damage to the embryo underneath: before air is allowed into the egg, the embryo will lie very close to the membrane.

6. Fill the cavity with CMF so that the embryo floats up to the level of the window.

7. After ensuring that there are no air bubbles in the syringe with Indian ink or in the needle, insert the needle under the vitelline membrane, tangentially, at a position as far away from the embryo proper as possible. Point towards and slightly below the embryo, and inject about 20-50 µl. It is important to minimise movement of the needle after penetrating the vitelline membrane, or the hole will be very large and yolk/ink will leak out. Introduce and withdraw the needle with one clean, decisive movement and do not stir the needle inside the yolk; only one attempt per egg!

8. Draw a shallow, continuous border of silicon grease around the window. This will contain a standing drop in which the operation will be done. Now fill this chamber with CMF saline until there is a standing drop, and adjust the fibre optic light to shine tangentially to the surface of the egg so that the embryo can be seen very clearly with minimal light intensity from the light source.

2.2.3. Grafting procedure

1. Break the vitelline membrane just over the region to be operated with a needle. The hole should be as small as possible. The segmental plates are the rod-like structures lying on either side of the neural tube at the tail end of the embryo, just behind the last somite. The portion to be rotated in this example is the most anterior half of the plate.

2. Replace the bubble of CMF with trypsin/CMF. Increase the magnification of the microscope as much as possible.

3. Operating in the drop of trypsin, use the micro-knife to make initially very shallow cuts in the ectoderm next to the neural tube in the region of the operation.

4. Gradually deepen the cuts using the knife blade more as a spatula, allowing penetration of the trypsin, than as a sharp cutting edge. Once the ectoderm has been penetrated with the tip of the blade, the trypsin does the rest. Find the lateral border of the segmental plate, and do the same there: a shallow cut in the ectoderm first, then separate the tissue gradually. In both cases, make sure you do not penetrate the endoderm (which is 1-cell thick) or ink will pour out. Finally, free the posterior end of the piece and loosen the graft.

5. Remove the piece of segmental plate with a Gilson micropipette set to 1-2 µl. Replace the bubble over the embryo with fresh CMF twice to remove the trypsin solution. Make a new bubble of CMF whilst obtaining the graft from the donor.

6. Turn to the donor embryo in the Sylgard dish. Replace almost all of the CMF in which it was submerged by trypsin solution. Repeat the trypsin wash and perform the dissection in this solution at room temperature.

7. Cut out an equivalent piece of segmental plate as the one removed from the host, using the same technique.

8. Pick up the graft with the Gilson. With the other hand, place the host under the microscope and, observing under low magnification, carefully place the graft into the CMF bubble over the embryo.

9. Use the knife and work at low magnification to manipulate the graft into the gap made by removal of the host piece of segmental plate.

10. When the graft is in position (approximately), very carefully remove the most of the fluid from above the embryo with a Pasteur pipette while watching under low power. If necessary, reposition the graft with a mounted needle.

11. Insert the 5 ml syringe with 21G needle into the original hole in the blunt part of the eggshell, vertically, and carefully withdraw 2.5-3 ml thin egg albumen. This will lower the operated embryo back to its original position. Be careful when you insert the needle, since the pressure could make the graft come out of its site.

12. Add 2-3 drops of antibiotic/antimycotic concentrate (away from the graft site).

13. Wipe the edges of the shell with tissue paper moistened lightly in 70% alcohol to remove the silicon grease.

14. Cut a piece of PVC tape about 6 cm long. Stretch it slightly, and then let it relax. Place it over the window, smoothing out any unevenness carefully so as to avoid breaking the shell or applying too much pressure on the window.

15. Keeping the egg on its side, place it (window down!) into an egg tray in a humidified incubator at 38EC. If you are worried about the graft falling out, it is a good idea to incubate the embryo with the window upwards until the next day, and then to turn it.

16. Incubate 1-3 days. Embryo survival 2 days after this operation should be 80-100%.


When the aim of the experiments is to investigate the mechanisms that set up the paraxial mesoderm, or when the fates and movements of presumptive somitic cells at or shortly after gastrulation are to be investigated, operations in the egg are very difficult. It is therefore generally necessary to resort to a method of whole embryo culture. For this, set up cultures of embryos at the desired stage (between HH stage 3 [mid-primitive streak], and HH stage 8 [4 somites]), following the protocols in other chapters in this volume (4, 5). Leave the embryo completely submerged in saline during the operation.

The region of the primitive streak embryo (stages 3-4) that contributes to the somites lies in the most anterior (cranial) 2/3 or so of the primitive streak and in the ectoderm to either side of this. By stage 4+-5, some of the somite progenitors have already left the streak and lie in the middle layer next to the anterior streak and gradually migrate cranially as more cells are added from the streak and node regions.

To operate on these cultured embryos, it is usually better to use fine mounted needles (entomological or sharpened tungsten) rather than knives. Trypsin is generally not necessary, but may be used (at 0.1% w/v) to facilitate separation of tissues if the specific manipulation desired turns out to be difficult due to adherence of the tissues to one another. Operations on the mesoderm are usually easier in the absence of trypsin because the middle layer readily separates into layers in the presence of enzyme. However, for each type of operation, the sequence in which the cuts are made is very important. In some regions of the embryo it is easier to begin with a medial cut and proceed laterally, while for some other regions an anterior cut, proceeding caudally is more effective. You should investigate the relative merits of different ways to dissect the tissue of interest before starting a set of experiments.

As described for operations on Hensen's node, it is important to avoid damaging the vitelline membrane at all costs. Any leakage of the albumen culture fluid to the inside of the ring will diminish survival and may cause the grafted tissue to fall out of its site.

After the operation, follow the steps described for Hensen's node grafts (5). Transfer the ring with the embryo to a plastic dish, seal it, place it in a humid chamber and incubate 38EC for 24-42 hours. If cultures are set up using large glass rings (about 30 mm; see 5) or other methods for extended culture (see 14), embryos operated at stages 3-6 should survive to stage 15 or even longer.


Embryos operated as described above can be studied by a variety of methods, according to the question being addressed. Fix the embryo in methanol (for most immunological detection procedures), Zenker's fixative (for Feulgen-Rossenbeck staining) or in 4% formaldehyde in PBS (for in situ hybridization and most other procedures). Details on these procedures may be found elsewhere in this volume and in ref. 15.

For embryos operated in ovo, it is easiest to crack the egg into a large Petri dish first, cut the membranes around the embryo with scissors, and then lift a corner of these membranes with fine forceps (as described above for "Preparation of the donor quail embryo"). Then immediately transfer this to a small dish with saline to clean off any adhering yolk. Finally transfer it to a Sylgard dish (see above) containing CMF and pin the embryo so as to straighten out the head and trunk, but avoiding the region close to the operation. Then remove the CMF and replace it with the fixative of choice. In this way, the embryos will be perfectly straight which will simplify subsequent histological sectioning, and they will also be more photogenic if stained as whole mounts.

For embryos operated in New culture (3-5), the glass ring should first be flooded with CMF, and the edges of the area opaca then detached from the vitelline membrane. Then pick up the embryo with a wide-mouthed pipette or with fine forceps (from the membranes!) and transfer it to a Sylgard dish for pinning and fixing as described above.


1. Keynes, R.J. & Stern, C.D. (1988) Mechanisms of vertebrate segmentation. Development 103, 413-429.

2. Tam, P.P. & Trainor, P.A. (1994). Specification and segmentation of the paraxial mesoderm. Anat. Embryol. 189, 275-305.

3. New, D.A.T. (1955). A new technique for the cultivation of the chick embryo in vitro. J. Embryol. Exp. Morph. 3, 326-331.

4. Hornbruch, A. (1996). New culture. (this volume)

5. Stern, C.D. (1996). Grafting Hensen's node. (this volume)

6. Hamburger, V. and Hamilton, H. L. (1951). A series of normal stages in the development of the chick embryo. J. Morph. 88, 49-92.

7. Selleck, M.A.J. & Stern, C.D. (1991) Fate mapping and cell lineage analysis of Hensen's node in the chick embryo. Development 112, 615-626.

8. Selleck, M.A.J. & Stern, C.D. (1992) Commitment of mesoderm cells in Hensen's node of the chick embryo to notochord and somite. Development 114, 403-415.

9. Stern, C.D. (1993). Transplantation in avian embryos. In: Essential Developmental Biology: A Practical Approach. (C.D. Stern and P.W.H. Holland, eds.) IRL Press at Oxford University Press, Oxford. pp. 111-117.

10. Lumsden, A. (1996). Grafting notochord. (this volume)

11. Lumsden, A. (1996). Grafting neural tube tissue. (this volume)

12. Tickle, C. (1996) Grafting AER and ZPA. (this volume)

13. Thorogood, P. & Hunt, P. (1996). Crest grafting, ablation and culture. (this volume)

14. Stern, C.D. (1993). Avian embryos. In: Essential Developmental Biology: A Practical Approach. (C.D. Stern and P.W.H. Holland, eds.) IRL Press at Oxford University Press, Oxford. pp. 45-54.

15. Stern, C.D. & Holland, P.W.H. (eds.) (1993). Essential Developmental Biology: A Practical Approach. IRL Press at Oxford University Press, Oxford. 333pp.

Figure legend

Figure 1. Schematic diagrams of embryos at about stage 11. the three regions of the somitic mesoderm can be seen along the axis of the embryo in the upper diagram: unsegmented mesoderm towards the bottom of the drawing, followed by epithelial somites (somite stages I through V or VI), followed more anteriorly by somites that have already split into dermomyotome and sclerotome (somite stages V/VI and higher). Although only 3 epithelial somites are shown, there are usually 5 or 6 of this type in embryos at this stage. The lower diagram represents a transverse section at the level of one of the epithelial somites.