An examination of non-formalin-based fixation ... - Wiley Online Library

12 downloads 2269 Views 486KB Size Report
Jun 17, 2005 - its lifelike attributes for an indefinite .... The quality of staining in the 2-hr .... for 2 hr at room temperature, the second row shows embryos fixed ...
DEVELOPMENTAL DYNAMICS 233:1464 –1469, 2005

TECHNIQUES

An Examination of Non-Formalin-Based Fixation Methods for Xenopus Embryos Anthony Acton,† Tia Harvey,† and Matthew W. Grow*

Despite the growing availability of non-formalin-based fixatives, the vast majority of researchers in developmental biology continue to fix embryos and tissue in 4% paraformaldehyde. This fixation method has proven useful for both immunohistochemistry and in situ hybridization, yet working with paraformaldehyde has distinct disadvantages in its toxicity and the short shelf life of prepared solutions. In a search for viable alternative fixatives, we have evaluated two non-formalin-based commercial products, FineFIX (Milestone Microwave Laboratory System) and NOTOXhisto威 (Scientific Device Laboratory). These products were tested side-by-side with a commonly used 4% paraformaldehyde solution (MEMPFA) on Xenopus laevis embryos and assayed using whole mount immunohistochemistry and whole mount in situ hybridization. The results indicate that NOTOXhisto威 can be used as a substitute for MEMPFA in both tested Xenopus protocols with no loss of sensitivity or tissue morphology. Developmental Dynamics 233: 1464 –1469, 2005. © 2005 Wiley-Liss, Inc. Key words: whole mount immunohistochemistry; antibody; in situ hybridization; fixation; fixative; non-formalin; paraformaldehyde; NOTOXhisto; FineFIX; Xenopus; embryo; MEMPFA; MEMFA Received 22 October 2004; Revised 5 February 2005; Accepted 26 February 2005

INTRODUCTION Tissue fixation is a vital step in tissue analysis, permitting the investigator to visualize the developmental stage at a distinct time point and maintain its lifelike attributes for an indefinite period. All subsequent stages in the experimental procedure hinge on proper fixation to interrupt detrimental postmortem cellular processes. The tissue can denature over time as a result of dehydrating agents or be subject to putrefaction or and/or autolysis. Due to diversity in reactive sites and rates of reactivity in individual tissue types, currently available fixatives may vary in their methods of

action. Fixatives may be classified into five major groupings including alcohols, aldehydes, mercurials, oxidizing agents, and picrates, based on their mechanism of action. Formaldehyde, or formalin-based fixatives, belongs to the aldehyde grouping and their mechanism of action involves cross-links between lysine residues that exist in the proteins (French and Edsall, 1945). These fixatives are renowned for their non-harmful effects on tissue and excellent tissue penetration. Conversely, formalin-based fixatives are highly toxic and have an intolerably short shelf life of a mere one to two weeks. Werner et al. presents a

review on this subject (Werner et al., 2000). Non-formalin-based fixatives are becoming more widely used as a safe alternative. Currently, however, the use of paraformaldehyde (usually in the form of MEMPFA/MEMFA) remains the standard protocol for the fixation of Xenopus embryos in preparation for in situ hybridization and immunohistochemistry. Here we examine in detail the performance of two commercially available non-formalinbased fixatives, NOTOXhisto威 (Scientific Device Laboratory) and FineFIX (Milestone Microwave Laboratory System) in both the detection of specific

Department of Biochemistry and Molecular Biology, Center for Medical Genomics, Indiana University School of Medicine, Indianapolis, Indiana † A.A. and T.H. contributed equally to this work. Grant sponsor: Indiana Genomics Initiative (INGEN); Grant sponsor: Indiana 21st Research and Technology Fund *Correspondence to: Matthew Grow, BRTC (L3) Room 226, 1345 W. 16th Street, Indianapolis, IN 46202. E-mail: [email protected] DOI 10.1002/dvdy.20448 Published online 17 June 2005 in Wiley InterScience (www.interscience.wiley.com).

© 2005 Wiley-Liss, Inc.

NON-FORMALIN-BASED FIXATION METHODS 1465

TABLE 1. A Comparison of Staining Intensity in Whole Xenopus Embryosa

MEMPFA NOTOXhisto FineFIX

12–101 2hr

12–101 48hr

IS ctrop 2hr

IS ctrop 48 hr

IS XNkx 2–5 2hr

IS XNkx 2–5 48hr

IS XNkx 2–5 2hr -PK

⫹⫹⫹ ⫹⫹⫹ ⫹⫹⫹

⫹⫹ ⫹⫹b ⫹⫹⫹

⫹⫹⫹ ⫹⫹⫹ ⫺

⫹ ⫹⫹ ⫺

⫹⫹⫹⫹ ⫹⫹⫹ ⫺

⫹⫹⫹ ⫹⫹ ⫺

⫹ ⫹ ⫺

a

This table summarizes the authors’ observations for all embryos in both the immunohistochemistry (12–101 antibody staining) and in situ hybridization (IS). The results are ranked from poor (⫹) to excellent (⫹⫹⫹⫹) quality based on the depth of color and size, along with the effects of embryo morphology. If no staining was present, the test group received no score (⫺). b High background staining present.

transcripts (using whole mount in situ hybridization) and protein detection (using immunohistochemistry) in Xenopus laevis embryos. We directly compare our results in side-by-side tests with MEMPFA, and discuss both the benefits and limitations of using these three fixatives. While all three fixatives worked equally well for immunohistochemistry, FineFIX failed in our in situ hybridization test. Additional tests further compare the effectiveness of NOTOXhisto威 versus MEMPFA. Our results indicate that NOTOXhisto威 offers a safe, cost-effective alternative to MEMPFA tissue fixation in Xenopus laevis.

RESULTS Whole Mount Immunohistochemistry To test the performance of MEMPFA, NOTOXhisto威, and FineFIX fixatives, we first conducted a test of immunohistochemistry. Embryos, fixed in test groups of twenty, were obtained and treated in identical conditions to avoid additional variables in the experiment. The study was carried out as articulated in the Experimental Procedures section using antibody 12-101 followed by alkaline phosphotase-conjugated anti-mouse secondary antibody and then an NBT/BCIP color development was completed to detect skeletal muscle staining. The results were then evaluated by multiple investigators to elucidate the following interpretation and summarized in Table 1 and Figure 1. The variation in staining is negligible between the embryos within all three fixative test groups. As seen in Figure 1, all treated embryos were found to retain integrity in contrast to

the in situ hybridization analysis and to yield equivalent antibody staining of skeletal muscle tissue. Sections of test embryos demonstrate that there is no significant difference in morphology between tested embryos, thus eliminating the concern that these non-formalin fixatives would generate the severe dehydration phenotypes described for other alcohol-based fixatives (Prento and Lyon, 1997). However, a certain amount of epidermal detachment and flaking was noted within the FineFIX-treated embryos. In the MEMPFA and NOTOXhisto威 fixative groups, the 48-hr room temperature incubation (attempted overfixation) generated inadequate staining compared to the optimal skeletal muscle staining produced within the 2-hr incubation group. Interestingly, the FineFIX fixative demonstrated better staining with the 48-hr incubation, although the noted epidermal detachment reduced the overall quality of the embryos. To prevent bias in the quality of staining within test groups, the results were not optimized in Figures 1 and 2. Rather the same amount of time was allowed for color development in each set of embryos.

Whole Mount In Situ Hybridization As described in Experimental Procedures, all embryos were siblings, with eggs and sperm obtained from the same source on the same day. A number of tests were conducted to ascertain the quality of performance of each fixative for the detection of mRNA transcripts via whole mount in situ hybridization. Two probes to detect the products of two different genes were synthesized: cardiac troponin I,

an abundant myocardial marker (Drysdale et al., 1994) and XNkx2-5, a homeobox transcription factor that is expressed at relatively low levels (Tonissen et al., 1994). Test groups of 20 embryos were assayed in each treatment, and the staining (or lack thereof) was fairly consistent among all embryos in each test group. Five representative embryos from each of the 15 test groups were photographed and are presented in Figure 2. Each test group was observed and the overall staining for each test group was rated individually by each of the authors. The observed results are summarized in Table 1. A clear failure in the staining of all embryos fixed with FineFIX is evident in Table 1 and the third column of Figure 2. These embryos also displayed epidermal detachment and tissue disintegration when treated with Proteinase K. To test for the requirement of Proteinase K treatment, test groups of embryos fixed for 2 hr at room temperature were subjected to XNkx2-5 detection using the standard Proteinase K step, while this step was withheld from another set of embryos. As previously reported that partial Proteinase K digestion is a necessary step in MEMPFA-fixed embryos, the results indicate that it is also essential for ample probe penetration and staining in NOTOXhisto威-fixed embryos. The quality of staining in the 2-hr MEMPFA and NOTOXhisto威-fixed embryos was nearly identical in both the cardiac troponin I- and XNkx2-5stained embryos. It should be noted that the color development was stopped at exactly the same time for

1466 ACTON ET AL.

tial preserved GFP staining with both fixatives.

MEMPFA Vs. NOTOXhisto: A Cost Comparison

Fig. 1. Antibody staining. All sibling Xenopus embryos were raised and fixed directly as in Experimental Procedures. Immunohistochemistry was performed to detect skeletal muscle using the 12-101 antibody on 20 whole embryos per test group (Kintner and Brockes, 1984). Very little variation was observed between embryos in a test group, and photographs of representative embryos are presented as follows: From left to right, each column shows embryos fixed in either MEMPFA, NOTOXhisto姞, or FineFIX. The first row shows embryos fixed in the respective fixatives for 2 hr at room temperature, the second row shows embryos fixed for 48 hr at room temperature, and the third row shows representative sections of paraffin-embedded embryos that were fixed for 2 hr. Note that the general morphology is preserved in these sections, although most of the FineFIX fixed embryos displayed some degree of epidermal detachment and flaking.

all test groups in Figure 2, and therefore the staining was not optimized for any particular set of embryos. There is a dramatic reduction in staining in both MEMPFA and NOTOXhisto威treated embryos fixed for 48 hr at room temperature. For the work presented in Figure 2, color development reactions were halted at the same time for all embryos. To demonstrate that equivalent staining can be obtained from either NOTOXhisto威 or MEMPFA fixation, another XNkx2-5 in situ hybridization was performed on a single batch of embryos in which the color development step is optimized. The color development for the 20 embryos treated with NOTOXhisto威 was allowed to proceed for 90 min whereas the 20 treated with MEMPFA were only developed for 60 min. The results were very consistent in each test group, with equivalent signal observed. Single, representa-

tive embryos from this test are shown in Figure 3.

GFP Fluorescence The use of Green Fluorescent Protein (GFP) has become commonplace among Xenopus researchers as a fluorescent reporter and lineage tracer. In order to determine if NOTOXhisto威 fixation would preserve GFP fluorescence, multiple 2-cell stage embryos were injected with synthetic GFP mRNA in one of the 2 blastomeres. Embryos were photographed prior to and following fixation with either NOTOXhisto威 or MEMPFA fixatives for (1 hr at room temperature). The embryos seen in Figure 4 are representative of the results obtained, and show that high, non-specific background fluorescence obscures most or any poten-

Regardless of the results from both the in situ hybridizations and immunohistochemistry, there is the very real, practical issue of reagent cost. Due to the promising performance of NOTOXhisto威, we performed a calculation of the per liter costs of both MEMPFA and NOTOXhisto威. Assuming purchases of 500-g quantities of paraformaldehyde and Magnesium Sulfate, as well as 100-g purchases of MOPS and EGTA, the cost of preparing a liter of MEMPFA is approximately $6.40. The current cost of one liter of NOTOXhisto威 is $7.57, a difference of $1.17 per liter. One must also take into account the additional value of the extended shelf life of NOTOXhisto威, yielding financial savings that may become quite significant. Institutional discounts were not applied in these calculations.

DISCUSSION Today, most researchers seeking to study the temporal and spatial expression of genes in whole tissue rely on formalin-based fixatives to preserve both nucleic acids and proteins in situ. Developmental biologists working with Xenopus embryos, with few exceptions, routinely fix embryos in MEMPFA (with 4% paraformaldehyde as the active ingredient) in preparation for whole mount in situ hybridization and (in some cases) immunohistochemistry. Despite the growing number of available non-formalin-based alternatives, there does not exist, of yet, a published account detailing the application of any of these alternatives for use on Xenopus embryos. In this work, we set out to determine if either of the two commercial fixatives tested could be used as a replacement for MEMPFA to preserve both proteins and mRNA in whole Xenopus embryos for detection using either immunohistochemistry or in situ hybridization, respectively. In our first experiment, we used antibody 12101 to specifically stain skeletal muscle (Kintner and Brockers, 1984). In

NON-FORMALIN-BASED FIXATION METHODS 1467

Fig. 3. Optimization of XNkx2-5 in situ hybridizations. Whole mount in situ hybridizations were performed to determine if low abundance transcripts (in this case XNkx2-5) could be detected to the same degree in NOTOXhisto姞fixed embryos as in MEMPFA-fixed embryos. The left embryo is representative of the staining observed in MEMPFA-fixed embryos after the final color development reaction was allowed to proceed for 60 min. The right embryo is representative of the staining in NOTOXhisto姞-fixed embryos where color development was allowed to proceed for 90 min. Note that the signal intensity is approximately the same, with minimal background staining.

Fig. 2. Whole mount in situ hybridization. As described in the Experimental Procedures section, whole mount in situ hybridization was performed to detect both cardiac troponin I (rows 1 and 2) and XNkx2-5 (rows 3–5). Embryos in columns 1–3 were fixed in MEMPFA, NOTOXhisto姞, and FineFIX, respectively. Embryos were either fixed for 2 hr at room temperature (rows 1, 3, and 4) or for 48 hr at room temperature (rows 2 and 5). The results were fairly consistent between the 20 embryos in each test group, and the five embryos shown are the most representative 5 embryos (showing the range of variance) in each of the 15 test groups. Note the degradation of FineFIX embryos treated with Proteinase K.

standard MEMPFA fixation, embryos are normally incubated 1–2 hr at room temperature, or overnight at 4°C (Sive

et al., 2000). All embryos treated for 2 hr in any of the three fixatives exhibited strong skeletal muscle staining.

Fig. 4. The effect of fixation on GFP-expressing embryos. The above embryos are representative of the results observed for more than 10 embryos per treatment. Embryos were injected in one blastomere at the 2-cell stage of development with 250 pg of synthetic GFP mRNA. The embryos were grown to stage 16 were they were viewed and GFP expression was photographed (left, “untreated” images) and individually tracked. One half of the embryos were then fixed with MEMPFA (top), while the other half was fixed with NOTOXhisto姞 (bottom). Both fixations were performed for 1 hr at room temperature. Individual embryos were then repositioned as closely as possible to their positions in original photographs, and their GFP expression recorded (right, “treated” images). Note that after fixation, specific GFP expression is diminished and/or background fluorescence obscures detection.

Over-fixation of tissues with paraformaldehyde occurs with extended incubation times resulting in high affinity bonding of the fixative with the lysine residues. Thus, increased cross-linking occurs and inhibits the access of probes (antibodies, antisense RNA, etc.) to their targets (Werner et al.,

1468 ACTON ET AL.

2000). To examine the effects of extended fixation, groups of embryos were allowed to incubate in the tested fixatives for 48 hr at room temperature. As expected, this resulted in a noticeable decrease in staining in the MEMPFA-fixed embryos. With both of the commercial fixatives, however, there was no observable decrease in signal, suggesting that over-fixation in either of these fixatives does not adversely affect antibody penetration or binding. Severe dehydration has been described as disrupting tissue morphology of tissues treated with alcohol-based fixatives (Prento and Lyon, 1997). For all fixatives tested here, however, tissue morphology was preserved in sections of embryos stained for skeletal muscle. Our second experiment was to test the use of these fixatives in a widely accepted protocol for whole mount in situ hybridization, using antisense RNA probes to detect either a high abundance (cardiac troponin I; Drysdale et al., 1994) or low abundance (XNkx2-5; Tonissen et al., 1994) transcript. Groups of embryos were incubated in the tested fixatives for either 2 hr at room temperature (a standard length of time, as previously described) or “overfixed” for 48 hr at room temperature. The results show that FineFIX-treated embryos performed very poorly in the detection of either transcript. Additionally, the embryos fixed in FineFIX were extremely sensitive to the Proteinase K treatment that is used in the in situ protocol to enhance probe penetration. NOTOXhisto威 performed much better in this test, with results similar to that of MEMPFA-treated embryos. In Figures 1 and 2, the test groups had no color development optimization in order to analyze fixative performance under identical conditions, whereas in Figure 3, the color development was optimized for NOTOXhisto威- and MEMPFAtreated embryos. The development time for NOTOXhisto威-treated embryos was a mere 30 min longer than that of the MEMPFA test group, demonstrating that the use of NOTOXhisto威 does not significantly increase the time required for equivalent results. It is interesting to note that while embryos treated with NOTOXhisto威 for 48 hr performed well in protein detection, such embryos still appear to be vulnerable to the effects of

over-fixation with respect to the in situ hybridizations. As seen in Figure 1, Proteinase K digestion continues to be an essential step in performing whole mount in situ hybridization on both NOTOXhisto威- and MEMPFA-fixed embryos. The GFP fluorescence study was conducted to further evaluate the span of use for the fixatives analyzed in this report. Both NOTOXhisto威 and MEMPFA fixation resulted in high background with little or no specific GFP expression visible. Based on this study, both NOTOXhisto威 and MEMPFA fixatives are susceptible to the same problems with regard to preserving GFP signal in Xenopus. This work was not designed to be a comprehensive study of all formalinbased fixative alternatives. Rather, we sought to test the hypothesis that at least one currently available commercial non-formalin fixative can be used as an alternative to paraformaldehyde fixation. We compared the non-formalin FineFIX and NOTOXhisto威 fixatives to the widely used formalin-based fixative, MEMPFA. The results demonstrate that NOTOXhisto威 can be substituted for MEMPFA in existing in situ and antibody detection protocols and, with optimization, can produce virtually equivalent results. In testing FineFIX in in situ hybridizations, we found it insufficient in preserving the morphological state of the embryo; thus, the specific transcripts were not preserved, resulting in little or no staining. It is important to note that we have not excluded additional untested fixatives or fixation methods. It may be of interest to researchers to further assess NOTOXhisto威 or additional fixatives on other organisms, such as whole zebrafish or mouse embryos. For the moment, however, our results suggest that researchers working with Xenopus embryos have a viable alternative to formalin-based fixatives that is cost-effective, environmentally safe, and has an indefinite shelf-life.

EXPERIMENTAL PROCEDURES Embryo Collection and Fixation Eggs were collected from one mature Xenopus laevis female frog injected

the prior night with 400U Human Chorionic Gonadotropin (HCG). Embryos were obtained through in vitro fertilization of several batches of eggs with minced Xenopus laevis testes. The developing embryos were then collected at stages ranging from 28 to 34 (Nieuwkoop and Faber, 1994), pooled and divided into 6 equal groups. Next, each group was directly fixed in MEMPFA for 2 hr, MEMPFA for 48 hr, NOTOXhisto威 for 2 hr, NOTOXhisto威 for 48 hr, FineFIX for 2 hr, or FineFIX for 48 hr. All fixations were performed at room temperature, gradually dehydrated into 100% Methanol, and then stored in 100% Methanol at -20°C. MEMPFA (also called MEMFA, stands for MOPS, EGTA, MgSO4, and Paraform Fixative) is 0.1M MOPS pH 7.4, 2 mM EGTA, 1 mM MgSO4, and 4% paraformaldehyde. NOTOXhisto威 fixative was obtained from Scientific Device Laboratory (http://www.scientificdevice.com); FineFIX (Milestone Microwave Laboratory System, http://www.milestonesrl. com) was obtained through Hacker Instruments and Industries, Inc.

Whole Mount Immunohistochemistry Test groups of 20 embryos were fixed in each fixative for either 2 or 48 hr (6 tubes total) and skeletal muscle was detected using the antibody 12-101 distributed by the Developmental Studies Hybridoma Bank at the University of Iowa (Kintner and Brockes, 1984). Embryos were processed using a slight modification of the Xenopus immunohistochemistry protocol from Early Development of Xenopus laevis: A Laboratory Manual (Sive et al., 2000). Embryos were rehydrated into TBT (Tris-based saline and 0.1% Triton X-100). A blocking reaction was run using TBT with 2 mg/ml BSA and 20% heat-treated goat serum. Embryos were incubated at 4°C overnight in 12-101 supernatant (neat). They were then washed in TBT followed by the same blocking reaction as before. An overnight incubation at 4°C was performed with an alkaline phosphotase-conjugated anti-mouse IGG secondary antibody (Pierce, ImmunoPure Antibody). Embryos were washed again in TBT and then an NBT/BCIP color development reaction was per-

NON-FORMALIN-BASED FIXATION METHODS 1469

formed with the reaction allowed to proceed for 12 hr at 4°C. Staining in all test groups was halted at the same time, refixed, viewed, photographed, and stored as stated below.

Whole-Mount In Situ Hybridization Unless otherwise noted, whole mount in situ hybridization on Xenopus laevis embryos was carried out using a method adapted from In Situ Hybridization: An Improved Whole-Mount Method for Xenopus Embryos (Harland, 1991). Antisense digoxigenin-labeled RNA probes were generated to detect transcripts from cardiac troponin I (Drysdale et al., 1994) and XNkx2-5 (Tonissen et al., 1994). Using the clones, pXTnIc linearized with NotI and pXNkxBSL linearized with BamHI. Probes were obtained from both clones through transcription using T7 RNA polymerase. Test groups of 20 embryos from each fixative treatment were assayed for either cardiac troponin I or XNkx2-5 expression using the protocol above. Embryos were rehydrated into a Tris Buffered Saline (TTw) and digested with proteinase K. Additionally, a test group for each 2-hr fixation was also processed with the XNkx2-5 probe with the Proteinase K digestion step removed. In the three groups without Proteinase K treatment, embryos went directly into the prehybridization steps after rehydration, while all others were acetylated and refixed with the corresponding fixative. All test groups were prehybridized at 60°C in hyb (⫹) buffer for 4 hr. Each tube was incubated at 60°C in its appropriate probe in hyb (⫹) buffer overnight. The optional RNAse treatment was omitted for all tubes. All tubes were then incubated overnight at 4°C with alkaline phosphotaseconjugated anti-digoxigenin antibody (Roche) and washed. An NBT/BCIP color development reaction was performed and allowed to proceed for 2 hr at room temperature. Staining in all test groups in Figures 1 and 2 was halted at the same time, refixed in

Bouin’s Fixative (LabChem Inc.), resuspended in 100% ethanol, and then viewed under a stereo Nikon microscope. Pictures were taken at 15x or 30x magnification of representative embryos individually as well as in groups. For the XNkx2-5 in situ used for Figure 3, the procedure was the same as above, except that the color development for test groups was allowed to proceed at 13°C and was halted at varying times dependent on equivalent signal. NOTOXhisto®-treated embryos were developed for 90 min and MEMPFA-treated embryos were developed for 60 min.

onto 25- ⫻ 75-mm glass slides, and the slides were dried at 37°C for 30 min. The paraffin was dissolved in Safeclear and either photographed immediately or mounted under coverslips using Permount mounting agent.

ACKNOWLEDGMENTS We greatly appreciate the support from our department and the Center for Medical Genomics (CMG, http:// cmg.iupui.edu), and the CMG Director, Dr. Howard Edenberg. We also thank Hacker Instruments and Industries, Inc., and Scientific Device Laboratory for providing complimentary product samples for testing.

GFP Fluorescence

REFERENCES

Synthetic mRNA for Green fluorescence protein (GFP) was generated using an SP6 mMessage mMachine reaction kit (Ambion, Inc.) from an Not1-linearized template from GFP CS2⫹plasmid (S65T modified) (Heim et al., 1995). A total of 250 pg was injected into multiple 2-cell stage embryos in one of the 2 blastomeres. Embryos were grown to stage 16 and were photographed prior to fixation using a Nikon fluorescent stereoscope at 60x magnification using a GFP Longpass filter (EX 470/40, DM 495, BA 500LP). Embryos that showed fluorescence in specific areas of the embryo were selected, and fixed for 1 hr at room temperature in either NOTOXhisto威 or MEMPFA fixatives (8 embryos each). The embryos were photographed a second time following fixation and evaluated for preservation of specific fluorescence.

Drysdale TA, Tonissen KF, Patterson KD, Crawford MJ, Krieg PA. 1994. Cardiac troponin I is a heart-specific marker in the Xenopus embryo: expression during abnormal heart morphogenesis. Dev Biol 165:432–441. French D, Edsall JT. 1945. The reaction of formaldehyde with amino acids and proteins. Adv Protein Chem 2:227–335. Harland RM. 1991. In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol 36: 685– 695. Heim R, Cubitt AB, Tsien RY. 1995. Improved green fluorescence. Nature 373: 663–664. Kintner CR, Brockes JP. 1984. Monoclonal antibodies identify blastemal cells derived from dedifferentiating limb regeneration. Nature 308:67–69. Nieuwkoop PD, Faber J. 1994. Normal table of Xenopus laevis (Daudin). A systematical and chronological survey of the development from the fertilized egg till the end of metamorphosis. New York: Garland Publishing, Inc. 260 p. Prento P, Lyon H. 1997. Commercial formalin substitutes for histopathology. Biotech Histochem 72:273–282. Sive H, Grainger RM, Harland RM. 2000. Early development of Xenopus laevis: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. 338 p. Tonissen KF, Drysdale TA, Lints TJ, Harvey RP, Krieg PA. 1994. XNkx-2.5, a Xenopus gene related to Nkx-2.5 and tinman: evidence for a conserved role in cardiac development. Dev Biol 162:325– 328. Werner M, Chott A, Fabiano A, Battifora H. 2000. Effect of formalin tissue fixation and processing on immunohistochemistry. Am J Surg Pathol 24:1016 –1019.

Sectioning After viewing the whole embryos, 2–3 representative embryos were selected, incubated in Safeclear (Fisher Scientific), perfused and then mounted in paraplast paraffin (Fisher Scientific). The embedded embryos were then cooled at ⫺20°C prior to sectioning to prevent curling of the sections. Each embryo was sectioned at 7-␮m thickness, sections were floated in water