Introduction

The stripe assay was originally designed by F. Bonhoeffer and co-workers1,2 in the late 1980s to analyze fundamental axonal guidance mechanisms governing the formation of the topographic map established in the chick retino-tectal system (for review, see ref. 3). Employing this assay, the presence of a graded distribution of repulsive axon guidance cues, present predominantly in posterior tectal membranes, was discovered. Subsequently, this assay was modified, permitting the identification of ephrin-As as the key components present in posterior membranes that are responsible for the repulsive properties4,5,6,7.

The stripe assay has since been used to uncover guidance cues and mechanisms in many principal neuronal projections, including, for example, the hippocampal, thalamic and olfactory systems8,9,10,11. The stripe assay was also applied to investigate neurite outgrowth and guidance behavior related to neuronal regeneration12,13,14,15 and branch formation16. Furthermore, the stripe assay has proven valuable in dissecting signaling cascades initiated downstream of axonal guidance receptors activated by ligands printed in a striped pattern12,13,14,15. In addition, the guidance properties of membrane-tethered molecules as well as the short-range action of secreted guidance molecules (such as Slits and Semaphorins) could be mimicked in vitro by this assay17,18. Also stripe assay experiments were conducted to unravel potential guidance properties of extracellular matrix components19,20 and of cell adhesion molecules of the IgG superfamily21.

Of note, the stripe assay was employed as an in vitro assay to investigate the navigation responses of both axonal growth cones toward guidance cues and of migrating cells4,22,23. With respect to the latter, neural crest cells22 and oligodendrocytes23,24, but also non-neuronal cells such as tumor cells, have been studied25.

The assay allows assessment of guidance activity of established molecules and also testing potential guidance properties of hitherto uncharacterized molecules such as RGM, the Wnt signaling inhibitor SFRP1 or the morphogen Sonic hedgehog26,27,28.

Besides the widely used stripe assay consisting of straight individual stripes, other geometric patterns—such as zig-zag patterns—have been employed29. These zig-zag stripes are valuable when analyzing cell behavior and signal-transduction events at substrate borders, where, for example, one part of a cell is adhering to a laminin substrate and the other part is contacting a substrate of different molecular nature30. Also, a novel technique allowing for microcontact printing of proteins on surfaces has been reported and offers a versatile tool to study the behavior of, for example, navigating growth cones in gradients of axon guidance cues31.

In the experimental setup, striped carpets are produced with the aid of specially manufactured silicon matrices to glass, plastic or nucleopore membrane surfaces.

These matrices consist of a channel system, which can be filled with a solution containing the guidance cue under investigation.

The original version of the assay employed stripes generated with crude membrane fractions of tissue (e.g., a certain brain area) or of cells transfected with the guidance molecule of interest. These crude membrane fractions are then sucked in a striped pattern onto polycarbonate nucleopore filters by application of vacuum.

The modified stripe assay allows one to examine the 'guidance properties' of purified molecules. Here, matrices are attached to either glass coverslips or plastic Petri dishes, which provide a solid and even surface to prepare the striped pattern. The choice of using either a coverslip or a Petri dish is at the experimenter's convenience. A detailed protocol is given for plastic Petri dishes, but glass coverslips have been used regularly and successfully by other investigators (see, e.g., ref. 18). After attaching the matrix to the Petri dish, the matrix channels are filled with the molecules of interest and thereby establish the set of first stripes (see Figs. 1,2,3). After incubation, the matrix is removed and the gaps between the first stripes are filled with a solution containing control molecules, thereby generating the set of second stripes. This experimental regimen results in alternating stripes consisting of guidance cue (“first stripes”) and control molecule (“second stripes”). To distinguish the set of first stripes from the neighboring second stripes, the first stripe solution contains fluorescently labeled antibodies or beads. To provide a general growth-promoting environment, laminin (or any other extracellular matrix component) is homogeneously applied to the carpets. After adding, for example, neurons on top of these carpets and allowing 1–2 days of culture time for neurite outgrowth/cell migration, stripe assay results can be evaluated. Preferential growth on the set of first (fluorescent) stripes demonstrates the presence of an attractive guidance cue on these stripes, whereas growth on the second stripes (containing the control molecule) is indicative of a repulsive activity of the guidance molecule placed on the first stripes. On top of that it should be kept in mind that a 'repulsive activity' as observed in the stripe assay could be due to a general non-permissive substrate property of the molecule under investigation. To demonstrate an active repulsive turning away of an individual growth cone from a local source of the guidance cue, the turning assay is widely employed within the axon guidance field (see ref. 32).

Figure 1: Original version of the stripe assay.
figure 1

(a) Experimental setup to produce alternating stripes of membrane fragments. Four suction filters mounted on a suction filter plate are connected to a vacuum pump via a non-return valve. Closing the hole (at black triangle) of the valve by inserting an Eppendorf tube leads to an increase of the pressure in the channel system (−800 to −1,000 mbar; 'high vacuum'). Removal of this tube decreases the pressure (−200 to −400 mbar; 'low vacuum'). (b) Detailed view of the suction filter plate illustrating the mounting of the matrices on the suction filters. Two filters are covered with white matrices (left side) to produce the set of first stripes. The other two filters, covered with blue matrices (right side), are used for generating the set of second stripes. Clips marked with white stars should always be closed when vacuum is to be generated. Clips on top are either closed (during changing nucleopore filters) or open (during generation of the striped pattern by applying vacuum or during washing steps). The dotted and dashed rectangles on the white and the blue matrix, respectively, indicate the position taken for the higher magnifications depicted in (c) and (e). (c) Higher magnification of a white matrix showing the channel system. Scale bar, 100 μm. (d) Electron microscopic picture of the white matrix channel system at even higher magnification. Scale bar, 350 μm. (e) Higher magnification of the blue matrix used to prepare the second stripes, depicting the innermost white mesh and the surrounding blue silicon base. Scale bar, 100 μm. (f) Even higher magnification of (e) highlighting the mesh-like structure of the innermost part of the blue matrix. Scale bar, 100 μm.

Figure 2: Modified version of the stripe assay.
figure 2

(a) Silicon matrix used to produce stripe patterns using purified proteins. Arrows point toward the inlet channel and the outlet hole. Scale bar, 0.5 cm. (b) Silicon matrices shown in (a) are pressed firmly with channels facing downwards (arrow) onto the surface of a Petri dish. (c) Margins of the stripes are marked on the bottom of the dish. (d) Illustration of method (a) to generate stripe patterns by injecting solution into the inlet channel with the help of a Hamilton syringe. The injected solution slowly fills and passes the channels and the excess volume exits at the outlet hole. During injection, the matrix is constantly pressed onto the dish surface. (e) Alternative method (b) of producing striped patterns by the use of suction. A volume of 100 μl of the first solution is pipetted onto the so-called 'outlet hole' and vacuum is applied at the 'inlet' channel. (f) After 30 min of incubation at 37 °C, excess solution is removed by washing with HBSS (method a or b). Excess liquid at the outlet hole is removed. (g) Starting at the upper right corner, the matrix should be lifted off the Petri dish and removed quickly. (h) A volume of 100 μl of second stripe solution is applied to the delineated area. Special care has to be taken not to touch or scrape the carpet surface with the pipette tip. Therefore, the solution is applied from one corner of the marked area. Subsequent washing steps and application of laminin or culture medium are performed in a similar way. The stripe pattern is encircled with a Pap-pen to reduce the volume of culture medium required.

Figure 3: Typical stripe patterns obtained.
figure 3

(a,b) Typical patterns of stripes obtained with the original method (a) (red stripes) and with the modified method (b) (green stripes). The circle in (b) delineates the position of an explant on top of this carpet. Scale bar, 100 μm. (c) Result of an original stripe assay experiment. Temporal RGC axons are repelled by posterior membranes (red) leading to growth on stripes derived from the anterior optic tectum. In contrast, nasal RGC axons are not repelled by posterior membranes thus growing randomly. Scale bar (ce), 100 μm. (d) Result of a modified stripe assay experiment. Alternating stripes of ephrinA5-Fc (red) and Fc alone were prepared. Both temporal and nasal RGC axon populations are repelled, thus growing on the non-labeled dark set of second stripes. (e) Result of a control experiment of the modified stripe assay. All stripes consist of Fc alone, resulting in random RGC axon outgrowth.

Herein we give a detailed outline of the stripe assay procedures using the chick retino-tectal projection. This protocol was adapted by various research groups for their experimental designs and, for these modifications, we would like to refer throughout the manuscript to the relevant primary research literature5,6,9,10,14,18,33. First we describe the guidance behavior of axons from chick retinal ganglion cells (RGCs) growing on carpets consisting of alternating membrane fractions derived from anterior and posterior chick optic tecta (original version, Step 1A of PROCEDURE). Second, a protocol describing the modified version of the stripe assay is provided (Step 1B of PROCEDURE), whereby alternating stripes of ephrinA5-Fc fusion proteins ('first stripes') and control stripes consisting of Fc alone ('second stripes') are prepared. Ephrin-As proteins trigger predominantly growth cone repulsion by activating EphA receptor tyrosine kinases on retinal axons34,35. Therefore, RGC axons are repelled by the set of first stripes containing ephrin-A5, leading to axonal growth congruent with the control set of second stripes.

Materials

Reagents

  • Animals/embryos: E6 and E9 white Leghorn chick embryos for preparation of RGC cultures and tectal membranes, respectively

    Caution

    All animal experiments must comply with national regulations.

  • Anti-human IgG, Fc-specific (Sigma, cat. no. I2136)

  • Anti-human IgG, Fc-specific Cy3-conjugated (Sigma, cat. no. C2571)

  • Aprotinin (Sigma, cat. no. A1153)

  • EphrinA5-Fc, human (R&D systems, cat. no. 374-EA)

  • Fc (Calbiochem, cat. no. 401104)

  • Fluorescent microspheres, 0.5 μm, for example red fluorescent (Molecular Probes, cat. no. F-8812)

  • Hank's balanced salt solution (HBSS; without Ca2+ and Mg2+(Invitrogen; cat. no. 14170-088)

  • Laminin (mouse, Invitrogen, cat. no. 23017015)

  • Leupeptin (Sigma, cat. no. L2884)

  • Methanol

  • PBS sterile (Invitrogen; cat. no. 14190), add 1 × penicillin/streptomycin

  • Pepstatin (Sigma, cat. no. P4265)

  • Spermidine (Sigma, cat. no. S2501)

  • SDS, 2% solution in H2O

  • Sucrose

Equipment

  • 60 mm bacterial Petri dishes (Greiner, cat. no. 628102)

  • Centrifuge, suitable for Falcon tube placement

  • Cell scraper

  • Falcon tubes, 15 and 50 ml

  • Hamilton syringe (Hamilton)

  • Laminar flow hood

  • Needle 26G, 0.45 μm

  • Silicon matrices (see EQUIPMENT SETUP)

  • Pap-pen (Invitrogen, Abcam)

    Critical

    Test Pap-pen for cytotoxicity before use.

  • Polycarbonate Nucleopore membranes, 0.1 μm (Whatman, cat. no. 112105). Cut nucleopore membrane into rectangles (approximately 1.5–1.7 cm × 1.2–1.4 cm). Label rectangles by cutting the upper left corner with shiny side facing upwards

  • Rotor tubes, 1.5 ml (Beckman)

  • Spectrophotometer (Pharmacia Ultraspec3000), 1 ml quartz curvette

  • Syringes, 1 ml

  • Ultracentrifuge (TL-100 Beckmann with TLS-55 rotor)

  • Equipment for suction apparatus: vaccum source (e.g., vacuum pump, laboratory bench vacuum); bottle; non-return valve fitting in bottle; tubing to connect bottle, plastic clips to close tubing (e.g., VWR), vacuum pump, suction filter plate and manometer

Reagent setup

  • Homogenization buffer (HB) 10 mM Tris/HCl, pH 7.4, 1 mM spermidine × 3 HCl and 1.5 mM CaCl2· 2H2O. Store at 4 °C.

  • Aprotinin Dissolve 25 mg in 4.2 ml autoclaved H2O, store aliquots at −20 °C; final concentration 200 U ml−1.

  • Leupeptin Dissolve 25 mg in 26.3 ml autoclaved H2O, store aliquots at −20 °C; final concentration 50 mM.

  • Pepstatin Dissolve 25 mg in 18 ml methanol, store aliquots at −20 °C; final concentration 2 mM.

  • 50% (w/w) sucrose solution Dissolve 50 g of sucrose in 50 g homogenization buffer, sterile-filter (0.4 μm) and store at −20 °C.

  • 5% (v/v) sucrose solution Dilute 50% sucrose solution 1:10 in homogenization buffer.

Equipment setup

  • Silicon matrices Silicon matrices and additional equipment for the stripe assay are presently distributed exclusively by the laboratory of Dr. Martin Bastmeyer (Zoologisches Institut, Universität Karlsruhe, Germany) (bastmeyer@bio.uka.de). For the original version of the stripe assay outlined in Step 1A of PROCEDURE, four white matrices to prepare the set of first stripes, four blue matrices to prepare the set of second stripes and the suction filter plate are required (see also Fig. 1). For the modified version of the stripe assay, outlined in Step 1B of PROCEDURE, the number of matrices (see Fig. 2a) required equals the number of carpets to be analyzed in one experiment. Regularly, we prepare up to 12 carpets in one experiment.

Procedure

  1. 1

    In the following steps we describe the procedure for the original version of the stripe assay in (A) and the modified version of the stripe assay in (B). For the stripe assay using purified proteins rather than crude membrane fractions1 (B), a modified protocol of the initially published work is employed4,36. In option B we describe a protocol preparing carpets of purified proteins belonging to the ephrin-A family of repulsive axon guidance molecules34,35. Ephrin-As are commercially available as fusion proteins (see MATERIALS section), containing the Fc part of the human IgG. Oligomerizing (so-called 'Clustering') of Ephrin-A-Fc with anti-Fc-specific IgG enhances the repulsion towards cells and growth cones by ephrin-A5. In principle, there are two options to produce stripes using the modified version of the stripe assay: injection of first stripe solution with a Hamilton syringe into the inlet channel (Fig. 2d) or by applying vacuum at the 'inlet channel' to suck solution through the channel system (Fig. 2e).

    1. A

      Stripe assays using crude membrane fractions (original version)

      1. i

        Preparations before starting. Boil (for 5 min) and store blue and white matrices in autoclaved H2O.

      2. ii

        Boil (for 5 min) and store precut nucleopore filters in autoclaved H2O.

      3. iii

        Coat the boiled nucleopore filters with 20 μg ml−1 laminin (in HBSS) and incubate for at least 1 h at 37 °C. Before use, wash with 1 × HBSS and store in PBS.

        Critical Step

        Thaw laminin slowly on ice (approximately 1 h) to prevent formation of aggregates.

      4. iv

        Prepare PBS with protease inhibitors (hereafter called PBS+) by adding 250 μl leupeptin, 100 μl aprotinin and 10 μl pepstatin to 10 ml PBS; store at 4 °C (prepare freshly for each experiment).

      5. v

        Prepare homogenization buffer with protease inhibitors (hereafter called HB+) by adding 125 μl leupeptin, 50 μl aprotinin and 5 μl pepstatin to 5 ml homogenization buffer; store at 4 °C (prepare freshly for each experiment).

      6. vi

        Thaw 50% sucrose solution, prepare 5% sucrose solution and store on ice.

      7. vii

        Sterilize 1.5 ml rotor tubes under UV light for 10 min.

      8. viii

        Preparation of crude membrane fractions. Prepare crude membrane fractions as a source of guidance molecules.

        As an example, in the following steps, we describe the preparation of membrane fractions from the anterior and the posterior thirds of six optic tecta of three E9 chick embryos. This amount of tissue represents approximately a total of 40–60 mg of cell material. The optimal amount of starting material has to be determined according to the individual experimental design. Alternatively, membrane fractions harvested of tissue culture cells transfected with a guidance cue of interest can be used (see ref. 5).

      9. ix

        Wipe the surface of chick egg with 70% ethanol and open the chick egg with blunt forceps.

      10. x

        Transfer the embryo into a Petri dish with HBSS.

      11. xi

        Remove the pia mater and dissect the two underlying tecta.

      12. xii

        Cut each tectum in three parts, discard the middle part and collect the anterior and the posterior thirds of both tecta in HB+ separately.

      13. xiii

        Wash the tectal thirds with 500 μl cold HB+.

      14. xiv

        Add 600 μl fresh HB+ and homogenize tissue several times using a 1 ml syringe with a 26G needle.

        Critical Step

        Avoid formation of air bubbles during homogenization. Homogenize slowly, that is, allow approximately 5–10 s for 600 μl homogenate to pass through syringe.

      15. xv

        Sucrose gradient centrifugation. This and the following steps describe sucrose gradient centrifugation, which will remove nuclei and cytoplasmic proteins and retain, for example, cytoplasmic and mitochondrial membrane fragments in the supernatant. First pipet 350 μl of 50% sucrose solution into a 1.5 ml rotor tube.

      16. xvi

        Layer slowly 150 μl of 5% sucrose solution on top.

      17. xvii

        Load a maximum of 800 μl homogenized membranes on top of the 5% sucrose cushion. The maximal amount of homogenate that can be separated represents the anterior/posterior thirds of three E9 embryos. When using transfected cells, for example HEK293 cells, the homogenate of four 10 cm dishes can be loaded5.

      18. xviii

        Centrifuge at 52,000g for 10 min at 4 °C using the TLS-55 rotor in a Beckmann TL-100 centrifuge. Meanwhile prepare Eppendorf tubes with 1 ml PBS+ and store on ice.

      19. xix

        After centrifugation, carefully remove the tubes from the rotor. Verify—by visual inspection—the presence of a white layer at an intermediate position in the rotor tube, which contains the crude membrane extract.

      20. xx

        Carefully transfer the white intermediate phase into the tube with 1 ml PBS+.

      21. xxi

        Centrifuge again to remove the remaining sucrose for 7 min at 13,000g at 4 °C.

      22. xxii

        Discard the supernatant and resuspend the pellet in 1 ml cold PBS+ with the syringe and 26G needle.

      23. xxiii

        Measurement of membrane fraction concentration. Add 20 μl of each sample to 980 μl of 2% (w/v) SDS solution (prepare duplicates). Pipette 20 μl PBS+ to 980 μl 2% SDS as reference. Determine OD at 220 nm in a spectrophotometer using a quartz curette.

      24. xxiv

        Multiply average OD220nm value by 3.33 to get the final OD220nm. 1 OD220nm = 700 μg ml−1. The average OD220nm value of three E9 tecta resuspended in 1 ml PBS+ (see above) is approximately 0.2.

        Pause point

        Aliquots of membrane fractions can be stored at −80 °C or in liquid nitrogen for several weeks, up to 1 month. Upon thawing, the activity should be measured again and those membranes should not be frozen again. We recommend to continue without delay, as membrane fractions lose some activity upon storage.

      25. xxv

        Dilute membrane fractions with cold PBS+ to an OD220nm of 0.1 and store on ice. Prepare 150 μl of membrane solution for preparing one set of first stripes and 300 μl of solution for preparation of one set of second stripes. As an estimate, adjusted membrane fractions with an OD220nm of 0.1 contain approximately 60–70 μg protein per ml as determined by Bradford assay.

      26. xxvi

        Label the first stripe solution with fluorescent beads by diluting 10 μl of beads in 500 μl PBS+. For 1 ml of membrane solution, add 50 μl of the diluted beads.

      27. xxvii

        Preparation of carpets. Set up the vacuum-sucking device according to Figure 1a. Use sterile blunt forceps to handle matrices and nucleopore membranes.

      28. xxviii

        Preparation before getting started with stripe preparation. Sterilize the suction filter plate by spraying with 70% EtOH; rinse individual suction filters with PBS (see Fig. 1b).

      29. xxix

        Check the vacuum with all four clips closed: at 'low' vacuum: −200 to −400 mbar (no Eppendorf tube in the bottle; see Fig. 1a); and at 'high' vacuum: −800 to −1,000 mbar (insert Eppendorf tube in the bottle). Closing or opening a hole in the bottle lid with the Eppendorf cup allows for switching between low and high vacuum.

      30. xxx

        Prepare 10 cm dishes with individual PBS+ drops to store each nucleopore membrane on top after the first stripes have been prepared. The PBS+ drop underneath the membrane will prevent the membranes from drying out.

      31. xxxi

        Prepare 35 mm Petri dishes with one drop of PBS+ in the middle to store carpets after preparation of second stripes.

      32. xxxii

        Preparation of first stripes. Cover each suction filter with one white matrix and rinse all matrices with 1 ml of PBS. Red lines of white matrix should be facing upward (see Fig. 1b,c).

      33. xxxiii

        Close all clips.

      34. xxxiv

        Add one nucleopore membrane on each of the four suction filters, with the marked corner of the membrane oriented to the upper left side (shiny surface faces downwards).

      35. xxxv

        Pipette 150 μl of first stripe solution on each of the four nucleopore membranes.

      36. xxxvi

        Open clips, and apply high vacuum by closing the hole of the vacuum bottle with an Eppendorf tube.

      37. xxxvii

        Start timer at −600 mbar and allow to suck for 90 s.

      38. xxxviii

        Remove excess liquid and stop high vacuum by removing the Eppendorf tube.

      39. xxxix

        Leave low vacuum until stripes are visible on filter surface.

      40. xl

        Immediately close clip and transfer all four nucleopore membranes on PBS+ drops in the Petri dish.

      41. xli

        The quality of the first set of stripes can be quickly checked by fluorescence microscopy.

      42. xlii

        Wash the white matrices before mounting the next nucleopore membranes by opening clips and rinsing with 1 ml of PBS.

        Critical Step

        Do not allow nucleopore membranes to dry out at any time.

      43. xliii

        Preparation of second stripes. Wash the suction filters three times with 1 ml of PBS by applying vacuum.

      44. xliv

        Add one blue matrix on each suction filter (in either orientation).

      45. xlv

        Lift the blue matrix slightly at the edge and suck with 1 ml PBS+ underneath.

        Critical Step

        Make sure blue matrices are lying completely flat on suction filters (no air bubbles underneath).

      46. xlvi

        Transfer nucleopore membranes (stored on PBS+ drop) onto suction filters.

      47. xlvii

        Immediately add 150 μl of second stripe solution and open clip; leave low vacuum until all four clips are opened.

      48. xlviii

        Apply high vacuum (insert Eppendorf tube) and suck for 90 s once −600 mbar is reached.

        Critical Step

        Do not allow nucleopore membranes to dry out at any time.

      49. xlix

        Rinse nucleopore membranes with 1 ml of PBS+.

      50. l

        Close clip and stop the high vacuum (remove Eppendorf tube).

      51. li

        Transfer processed nucleopore membranes to individual 35 mm Petri dishes (with a drop of PBS+ at the bottom).

      52. lii

        Add one drop of tissue culture medium onto the nucleopore membrane.

        Pause point

        Store dishes at 37 °C and use on the same day (i.e., as soon as cells are ready to be cultured on carpets).

    2. B

      Stripe assays (modified version) using purified proteins

      1. i

        Preparations before starting experiment. Remove dust from matrix stripe surface with transparent tape. Boil matrices for 5–10 min in autoclaved H2O and dry overnight with striped pattern facing upwards in a laminar flow hood.

        Critical Step

        Matrices should be completely dry before being used.

      2. ii

        Sterilize 60 mm bacterial dishes under UV for 5–10 min.

      3. iii

        Thaw laminin aliquot slowly on ice.

        Critical Step

        This step has to be performed slowly on ice to prevent formation of aggregates (approximately 1 h).

      4. iv

        Rinse Hamilton syringe with ethanol, followed by autoclaved H2O and HBSS between individual experiments. Sterilize outside by UV.

      5. v

        Preparation of solutions and Petri dishes. For every carpet, prepare 100 μl solution in HBSS for each of the first and second stripes. For preparation of the first stripes, two solutions are required: one ephrin-A5-containing solution for the experimental situation (1. stripe: ephrin-A5-Fc versus 2. stripe: Fc) and one Fc-containing solution for the control setting (1. stripe: Fc versus 2. stripe: Fc). For the second stripes, only one solution containing Fc alone has to be prepared (see also controls in Box 1). First stripe solution: ephrin-A5-Fc or Fc, 10 μg ml−1. Anti-human-IgG Fc-specific Cy3 conjugated, 2.5 μg ml−1. Second stripe solution: Fc, 10 μg ml−1. Anti-human-IgG Fc-specific, 2.5 μg ml−1. For alternative concentrations, see refs. 33,36. The approximate active concentration when testing novel proteins is recommended to be within a range of 1–30 mg ml−1. Spin down anti-human-IgG, Fc-specific antibodies (Cy3-conjugated and unconjugated) before use (1–2 min, 13,000g, 4 °C).

        Critical Step

        Without centrifugation, antibodies might produce aggregates resulting in unspecific guidance responses.

      6. vi

        Preincubate first and second stripe solutions for 30 min under moderate agitation to allow for oligomerization of ephrinA5-Fc; keep solutions with Cy3-conjugated antibodies in the dark.

      7. vii

        Prepare Petri dishes with matrices meanwhile by adding one matrix with channels facing downwards in the middle of a 60 mm bacterial Petri dish (see Fig. 2b). Label margins of stripes using a marker pen at the bottom of the dish (Fig. 2c).

        Critical Step

        Remove any air bubbles between matrix and dish surface, as this may cause leakage of solution.

      8. viii

        Preparation of the first set of stripes. Prepare the first set of stripes by applying first stripe solution to the matrices. There are two possible methods: in the first (injection with a Hamilton syringe), insert the bent tip of a Hamilton syringe into the inlet channel of one matrix and inject slowly 100 ml of the first stripe solution (Fig. 2d). The solution should slowly pass the stripe channels and leave the matrix opposite to the inlet channel. In the second method (using a vacuum), add 100 μl of first stripe solution on outlet hole and suck solution through the channels by applying vacuum at inlet channel (Fig. 2e). It is critical to leave a small drop covering the outlet channel to prevent drying out.

      9. ix

        Once all matrices of all dishes have been injected with first stripe solution, incubate all dishes for 30 min at 37 °C in a moist chamber to prevent drying out of stripes.

      10. x

        Wash by slowly rinsing the matrices with 500 μl HBSS employing either the Hamilton syringe or the vacuum method (as described in Step viii).

      11. xi

        Remove excess HBSS and matrix from the Petri dish by lifting the matrix starting at the upper right corner opposite to the injection channel (Fig. 2f,g).

      12. xii

        Preparation of the second set of stripes. Immediately after removing the matrix, add 100 μl of second stripe solution to the previously labeled area of stripes (labeled in Step 1B(vii), also see Fig. 2h).

        Critical Step

        After removing the matrix of one dish, immediately add second stripe solution to prevent drying out of first stripes. Do one dish at a time.

      13. xiii

        You may find it helpful to encircle the stripes with a Pap-pen (approximately 2 cm in diameter; Fig. 2h) to reduce the total volume of cell culture medium needed; however, this is an optional step.

      14. xiv

        Incubate all dishes for 30 min at 37 °C.

      15. xv

        Remove second stripe solution as much as possible but not completely to prevent drying out of the carpet and wash once with HBSS.

      16. xvi

        Incubate stripes with 100 μl laminin in HBSS for 2–3 h at 37 °C (20 μg ml−1). Laminin provides an overall outgrowth-promoting substrate, which is necessary, particularly when testing repulsive molecules; laminin incubation might be skipped depending on the molecule under investigation. Also other extracellular matrix components such as fibronectin or collagen can be tried. During laminin incubation, prepare cells to be examined. Note that this step is optional and might be omitted while testing novel proteins for their guidance activity.

      17. xvii

        Wash stripes once with HBSS.

      18. xviii

        Add 500 μl culture medium and keep dishes in the incubator until cells are ready to be plated.

    Critical Step

    All steps, if not indicated otherwise, are performed in a laminar flow hood.

Plating of cells and explants on carpets

  1. 1

    For plating dissociated single cells (e.g., mouse hippocampal neurons), spread 104 cells per stripe area. Mouse postnatal hippocampal neurons can be prepared according to a previously published protocol37. When using explants (e.g., dorsal root ganglia, vomeronasal organ, Xenopus or zebrafish retina), distribute 10–20 explants on one carpet. To allow radial outgrowth of nerve fibers from the explant, sharp tungsten needles should be employed to cut tissue into squares or rectangles (average diameter 300–500 mm). Alternatively, as in the case of chick RGC explants (see ref. 2), E6 retinae can be flat-mounted on nitrocellulose filters and cut into 200–300 mm wide strips using a tissue chopper. Two of these retinal explants can be arranged on one carpet.

  2. 2

    Incubate cultures for 24–48 h. If desired, perform standard immunocytochemistry employing, for example, a tubulin-specific antiserum or fluorescent labeling of RGCs with DiI2.

    Troubleshooting

Troubleshooting

Troubleshooting advice can be found in Table 1.

Table 1 Troubleshooting table.

Timing

To prepare 10–12 carpets either by the original or by the modified version of the stripe assay, it takes approximately 4–5 h.

Anticipated results

Images depicting typical anticipated results are given in Figure 3. Figure 3a shows the typical pattern of alternating stripes derived from membrane fractions of the anterior and the posterior chicken optic tectum, respectively. Posterior membranes are labeled with rhodamin-coupled beads. Figure 3b shows alternating stripes consisting of purified ephrin-A5-Fc (labeled with FITC-conjugated anti-human IgG, Fc-specific antibodies) versus Fc alone (dark stripes). Note that stripe borders using the modified version of stripe assay appear sharper compared to the original version (see Fig. 3a). The circle indicates the position and size of an explant cultured on top of the carpet. Figure 3c shows a stripe assay experiment using alternating stripes of anterior and posterior tectal membranes (original version), cultured with RGC explants (labeled in green). Axons of the temporal retina are strongly repelled by cues present in the posterior membranes (labeled in red). Nasal axons growing in a random fashion serve as an internal control in this type of experiment. Figure 3d shows a stripe assay experiment using purified ephrin-A5-Fc (modified version). Ephrin-A5-Fc-containing stripes (in red) are alternating with Fc-only-containing stripes (not marked). In this setup, both temporal and nasal RGC axons (in green) are repelled by ephrin-A5-Fc, thus growing preferentially on Fc-containing stripes. Note that owing to fasciculation, not the entire stripe width is occupied by RGC axons.

Figure 3e shows a control stripe assay experiment using carpets with both stripes consisting of Fc alone. This important control—to be included in every stripe assay experiment—reveals unbiased random outgrowth of axons.