DEVELOPMENTAL DYNAMICS 207:204-214 (1996) Expression of Multiple CD44 Isoforms in the Apical Ectodermal Ridge of the Embryonic Mouse Limb QIN YU,NICHOLAS GRAIMMATIKAKIS, AND BRYAN P. TOOLE Department of Anatomy and Cellular Biology, Tufts University School of Medicine, Boston, Massachusetts 02111 ABSTRACT Previous immunohistochemical studies have shown that CD44 is highly enriched within the apical ectodermal ridge of the developing limb (Wheatley et al. [19931 Development 119 295-3061, but the particular isoforms of CD44 were not identified. We show here that CD44s (standard or “hemopoietic” isoform) and several CD44 variants, especially V3-Vl0, VPV10, and V6V10, are concentrated in the apical ectodermal ridge in the early mouse limb. Since CD44s is a major cell surface receptor for hyaluronan, we compared its localization with that of hyaluronan. In the early limb bud, hyaluronan is distributed throughout the mesoderm but is absent from all regions of the ectoderm. Hyaluronan is especially enriched in the basement membrane separating ectoderm and mesoderm, except beneath the apical ectodermal ridge where it is absent. Since CD44s is a known endocytic receptor for hyaluronan, its presence in ridge ectoderm could lead to degradation of hyaluronan destined for the neighboring region of basement membrane, thus facilitating interaction of the ridge with underlying mesoderm. The CD44 (V3-VlO) isoform found in the ridge is expressed elsewhere as a proteoglycan with heparan sulfate chains that bind fibroblast growth factors. Since fibroblast growth factors are present in the ridge and are essential for limb morphogenesis, CD44 (V3-V10) is likely to act as a cofactor or modulator in the growth-promotingaction or maintenance of the ridge. o ISSS Wiley-Liss, Inc. Key words: Mouse limb development, CD44 isoforms, Hyaluronan with cytoskeletal proteins (Lacy and Underhill, 1987; Bourguignon et al., 1993; Tsukita et al., 1994) and contains serine residues that can be phosphorylated (Neame et al., 1992). Exons 6-15 of CD44 (usually termed V1-V10) are alternatively spliced into the membrane-proximal region to give numerous additional isoforms (Screaton et al., 1993; Sherman et al., 1994). CD44 is also extensively glycosylated, with both oligosaccharide and glycosaminoglycan side chains, thus generating further diversity among the various CD44 isoforms. Although it is clear that hyaluronan is a major ligand for CD44s (Aruffo et al., 1990; Culty et al., 19901, CD44 variants do not usually bind hyaluronan as avidly as CD44s (Jackson et al., 1995). However, interactions within the plasma membrane can influence binding to hyaluronan in a positive or negative fashion (He et al., 1992; Lesley et al., 1992, 1993). As yet, other ligands for most CD44 variants have not been clearly established, except that proteoglycan forms of CD44 bind extracellular matrix proteins (Jalkanen and Jalkanen, 1992; Faassen et al., 1992) and growth factors (Tanaka et al., 1993; Bennett et al., 1995) via their glycosaminoglycan side chains. Of special interest is the recent finding that CD44 variants incorporating the V3 exon are preferentially conjugated with heparan sulfate side chains that bind basic fibroblast growth factor (FGF-2) and other heparinbinding growth factors (Jackson et al., 1995;Bennett et al., 1995). CD44 isoforms containing variant exons are commonly found in normal adult and fetal epithelia (Terpe et al., 1994; Mackey et al., 19941, in several types of leukocytes (Arch et al., 1992; Mackay et al., 19941, and in metastatic cancer cells (Gunthert et al., 1991; Rudy et al., 1993). Immunohistochemistry has been used to document the presence of CD44 during specific stages of development of a variety of tissues and organs (Underhill, 1993; Wheatley et al., 1993; Pavasant et al., 1994). One of the major conclusions to emerge from the latter work, together with several cellular and biochemical studies (Orkin et al., 1982; Culty et al., 1992, Hua et al., 1993; Underhill et al., 19931, is that CD44s mediates internalization of hyaluronan en route to in- INTRODUCTION CD44 is a widely distributed integral cell surface glycoprotein expressed in many adult and embryonic tissues (Kennel et al., 1993; Wheatley et al., 1993; Lesley et al., 1993). There is a single gene for CD44 which, in the mouse, contains 20 exons. Exons 1-5, 16-18, and 20 give rise to CD44s (standard or “hemopoietic” form of CD44), the predominant isoform present in most tissues. This isoform contains an extracellular domain with an N-terminal hyaluronanbinding site and a membrane-proximal region, a transReceived January 25, 1996;accepted May 22, 1996. membrane domain, and a C-terminal cytoplasmic doAddress reprint requestslcorrespondence to Bryan P. Toole, Departmain (Stamenkovic et al., 1989; Goldstein et al., 1989; ment of Anatomy and Cellular Biology, Tufts University School of Zhou et al., 1989). The cytoplasmic domain interacts Medicine, Boston, MA 02111. 0 1996 WILEY-LISS, INC. CD44 ISOFORMS IN LIMB DEVELOPMENT 205 product obtained using RNA from the rest of the limb tissues was -110 bp, corresponding to CD44s (Fig. lC, lane 3). The PCR products detected above were cloned into the pCRII vector, and then plasmids were prepared and digested with EcoR1. The inserts were subjected to agarose electrophoresis to confirm their size and purity (Fig. 1D) and fully sequenced. The sequences were then compared with published mouse genomic sequences for the various CD44 isoforms (Screaton et al., 1993). The sequences of the 110-bp products obtained from RNA derived from AER-enriched tissue or from the remaining limb tissues corresponded in both cases to the expected regions of exons 5 and 16 of CD44, thus confirming the presence of CD44s in these tissues. The two larger products of -750 bp and -1,000 bp, detected only in the AER-enriched tissue, were found to contain the same partial exon 5 and 16 sequences as above plus the entire V6-V10 and V4-V10 groups of exons, respectively. In addition we cloned and fully sequenced several minor products from the AER-enriched tissue; these isoforms contained exons V4N6N9-Vl0, V4l V8-Vl0, V7-VIO, V8-VIO (E-form), V9, and V9-10 (Fig. 2). To verify the above results and to obtain additional information regarding the types of CD44 variants present in the AER region of the limb, we used additional RT-PCR strategies. First, we used sense primer 2 from the experiments above (Fig. lB), together with antisense primers corresponding to portions of exons V3 through V10, as described in Experimental Procedures and Table 1. In this series of experiments two major products were obtained (Fig. 3A); these products corresponded in size to the CD44 variants containing exons V4-V10 and V6-V10 and were verified as such by partial sequencing, thus confirming our results RESULTS above. Second, we used primer 1 from above (Fig. 1B) Identification of Several CD44 Isoforms in the with sense primers from exons V1 through V10. An Distal Region of the Early Limb Bud additional major amplified product was obtained (Fig. Approximately 100 E l 0 mouse limbs were dissected 3B, lane 4)) which was sequenced completely and ideninto two regions, one of which contained the AER plus tified as corresponding to the CD44 variant containing a small amount of adherent mesoderm; the other re- exons V3-V10. In both series, several other minor gion included the rest of the limb (see Fig. 1A). Total variants were detected and confirmed, by partial seRNA was then prepared from separate pools of these quencing, to be the same as those found by the first limb regions, and reverse transcription-polymerase strategy. Also, small amounts of CD44(V10),and CD44 chain reaction (RT-PCR)was performed as described in (V2-V3) were found. In summary, CD44 isoforms expressed in AER-enExperimental Procedures, using primers 1and 2 which flank the site of insertion of variant exons in CD44 riched tissue from E l 0 mouse limbs include CD44s, (Fig. 1B). If CD44s is the only isoform present in these CD44 (V3-V10), CD44 (V4-V10) and CD44 (V6-VlO). tissues, the RT-PCR product would be 110 bp in size Several other minor variants are also present. The maand would include portions of exon 5 and exon 16 of jor isoform expressed in the rest of the limb is CD44s. CD44. If isoforms other than CD44s are present, larger products would be obtained. As shown in Figure lC, Localization of CD44 Isoform mRNAs in the lane 2, three significant amplified products were ob- AER by In Situ Hybridization The dissected regions of the E l 0 limbs used above for tained using RNA prepared from the AER-enriched tissue. One of these was -110 bp and thus corresponded RT-PCR were cross-contaminated to a small extent. to CD44s; however, the other two products were larger, Thus, for example, it is possible that the variants de-750 bp and -1,000 bp. Several minor bands were also tected in the AER-enriched tissue could have been deobtained between 200 and 700 bp. The only significant rived from a small amount of contaminacing subecto- tracellular degradation. A second conclusion is that CD44 is often present at sites of epithelial-mesenchyma1 interactions that are known to be important in morphogenesis (Underhill, 1993; Wheatley et al., 1993). However, the CD44 isoforms present in these developing tissues have not been identified or specifically localized, although preliminary data obtained by size estimation (Wheatley et al., 1993) and polymerase chain reaction (PCR) (Ruiz et al., 1995) indicate that many isoforms are likely to be present during morphogenesis of a number of tissues and organs. The studies of Wheatley et al. (1993) demonstrated a striking localization of CD44 to the apical ectodermal ridge (AER) of the mouse limb bud. The AER is known to interact with distal limb mesoderm and thereby control growth and shape during early stages of limb bud morphogenesis (Saunders, 1948; Summerbell et al., 1973). FGF-2, FGF-4, and FGF-8 are present in the AER and appear to mediate at least part of the function of the AER (Niswander et al., 1993; Fallon et al., 1994; Crossley and Martin, 1995; Ochiya et al., 1995). Since V3-containing isoforms of CD44 have been shown to bear heparan sulfate side chains that bind FGF (Bennett et al., 1995), it would be of particular interest to know whether these forms are present in the AER. In addition, whereas there is considerable evidence that hyaluronan plays an important role in the chick embryo limb (Toole et al., 1991), a recent study failed to detect hyaluronan in the developing mouse limb (Fenderson et al., 1993). Since CD44s is a likely component of the AER (Wheatley et al., 1993), the relative distribution of hyaluronan and CD44s is also of significant interest. Because of the above issues, the current study was conducted to establish the nature of CD44 isoforms during morphogenesis of the mouse embryo limb. 206 YU ET AL. fl. 1 Dissection o f E l 0 limb b u d Mesoderm \- beneath AER Plane o f dissectlon B. CD44 gene: v6 v7 V8 V9 V10 Fig. 1. Identification of CD44 isoforms in AER-enriched tissue by RfPCR. A: Diagram of E l 0 mouse limb with the AER shown in longitudinal section. The AER-enriched region was dissected along the thick line, thus separating the AER and a small amount of attached mesoderm from the rest of the limb. Total RNA was then prepared from these two areas for RT-PCR. B: Diagram of CD44 gene showing the exons of CD44s (sl-s9) and of the variant isoforms (VI-VIO), as well as the locations of antisense primer 1 and sense primer 2. C: Agarose electrophoresis of RTPCR products. Lane 1: 100-bp DNA ladder. Lane 2: RT-PCR using total s6 s7 s8 $9 RNA from the AER-enriched tissue. Lane 3: RT-PCR using total RNA from the remaining tissues of the limb. Arrowheads indicate the major bands: CD44 (V4-V10), CD44 (V6-VIO), and CD44s in lane 2 and CD44s in lane 3. 0 : pCRll clones of RT-PCR products. Lane 1: 100-bp DNA ladder. Lanes 2-1 1: pCRll clones containing inserls corresponding to the RT-PCR products from CD44s, CD44 (V4-V10), CD44 (V6-V10), CD44 (V4N6N9-V10), CD44 (V4N8-V10), CD44 (V7-V10), CD44 (VS-VlO), CD44 (V9), CD44 (V9-VIO), and CD44 (V3-V10), respectively. CD44 ISOFORMS IN LIMB DEVELOPMENT 207 CD44 isoform RT-PCR products $5 $6 C044s $5 U4 U5 U6 U7 U8 U 9 U l O s 6 ‘4 < I U4-UlO s 5 U 6 U 7 US U9 U 1 0 s 6 U6-U10 6 5 U 4 U 6 U9U10a6 t 5 U 4 U 8 U9UtOs6 $5 U7 U8 U9 U10 $6 U7-UlO s 5 U S U 9 UlOs6 U8-UlO $5 U 9 s 6 u9 $5 U9 U10 $ 6 DsEzl U9-UI 0 U3 u4 U5 U6 U7 U8 U 9 U 1 0 s 6 UJ-U1D Fig. 2. CD44 isoforms identified by RT-PCR, pCRll cloning, and sequencing. lsoforms were identified as described in Figure 1, and comparison of sequences was obtained with published CD44 genomic sequences (Screaton et al., 1993). The major RT-PCR products identified corresponded to CD44s. CD44 (V4-V10), and CD44 (V6-VlO); other minor products identified were V4N6N9-VI0, V4NB-VI 0, V7-VI0, V8-VIO (E form), VQ, and V9-VIO. An additional prominent product, CD44 (V3-VIO), was identified as described in Figure 3. dermal mesoderm rather than the AER itself. Consequently we determined the precise location of these variants by in situ hybridization using several digoxigenin-labeled antisense riboprobes. Probe H3-5 (410 bp) is derived from the 5’ common region a€0 4 4 and would detect all isoforms. Probe V4-5 (230 bp) contains exons V4 and V5 and would detect variants Fig. 3. Confirmation of identities of major CD44 variants in AER-enriched tissue. A: Sense primer 2 and antisense primers from exon VIO, V9, V8, V7, V6, V5. V4, or V3 were used in RT-PCR (lanes 2-9, respectively); lane 1, 100-bp DNA ladder. The sizes obtained for the two major products are those expected for CD44 variants, V4-VIO and V6-VIO (arrowheads in lanes 2-6); the arrowheads in lanes 7 and 8 indicate the product expected for V4-V10; the arrowhead in lane 9 indicates that expected for V3-V10. 8: Antisense primer 1 and sense primers from exon V1, V2, V3, V4,V5, V6,V7, V8, V8, or V10 were used (lanes S11, respectively): lane 1, DNA ladder. The size obtained for the major product in lane 4 (arrowhead) is that expected for CD44 variant, V3-V10. The identities of the variants were conflrmed by sequencing. CD44 (V3-V10) and CD44 (V4-VlO) but not CD44s, CD44 (V6-V10), or several of the other minor variants. Probe V4-7 (472 bp), probe V6-7 (242 bp), probe V6-10 (406 bp), and probe V8-VIO (398 bp) would detect most variants but not CD44s. As controls, sense probes corresponding to all of the above antisense probes were used. Figure 4 shows the results obtained with some of the above probes for the El0 limb bud. All of the antisense probes gave a strong signal €or mRNA in the AER, although the short probes (V4-5 and V6-7) gave slightly less intense signals than the longer probes. 208 YU ET AL. TABLE 1. RT-PCR Primed Antisense primers 1 (exon 16): 5'-GATCCATGAGTCACAGTGCG-3' ~ 1 0 5'-CTGGTAAGGAGCCATCAACA-3' : v9- 5'-eTAGATGGCAGAATAGAAG-3 v8: 5'-CTGTTCAAGTC?TCCACCAA-3' ~ 7 5'-TGC'M"I'CTGTTTGATGACC-3' : ~ 65'-CAGT"GTCCCWCTGTCACA-3' : v5: 5'-TTGTGCTTGTAGCATGTGGG-3' ~ 45'-AACCCGTGGAGTACITGCAA-3' : ~ 35'-TGGTACTGGAGATAAAATCT-3' : Sense primers 2 (exon 5): 5'-GCCTACTGGAGATCAGGATG-3' vI: 5'-TTGCCTCAACTGTGCACTCA-3' vZ:5'-TGATGACCACCCCTGAAACA-3' v5I 5'-A?AGACAGAATCAGCACCAG-3' ~ 65'4TCCTAATAGTACAGCAGAA-3' : ~ 7 5'-CTTCGGCCCACAACAACCAT-3' : v8: 5'-ATACAGACTCCAGTCATAGT-3' v9: 5'-CACAGAGTCATTCTCAGAAC-3' ~10: 5'-CGTTAATGTTGATGGCTCCT-3' T h e sequences were derived from Screaton et al. (1993) (GenbanWEMBL accession number L13611). Very low signals were obtained in the remaining ectoderm and in the mesoderm in all cases (Fig. 4A-C). NO signal was obtained in any region of the limb with any of the sense probes (e.g., see Fig. 4D). These results, along with the results of RT-PCR, indicate that all of the variant CD44 isoforms are virtually restricted to the AER. Although CD44s was detected in the mesoderm by RT-PCR, the in situ signal observed with all of the probes was diffuse and close to background level. This continued to be the case at later developmental stages prior to differentiation. As differentiation progressed in the limb, CD44s was detected by RT-PCR and in situ hybridization in putative osteoprogenitor cells within the erosion zone of developing bone (also see Pavasant et al., 1994) and in the perichondrium. The E isoform, CD44 (VS-VlO), was also present in the perichondrium. No other variants were detected in these structures (data not shown). Western Blot Analysis of CD44 in Limb Tissues Although our in situ results showed that the concentration of CD44 in the mesoderm and non-AER ectoderm is low relative to the AER, our RT-PCR results indicated that some CD44s is present outside the AER. Unfortunately, it is not possible to design an in situ probe specific for CD44s, and therefore we cannot directly demonstrate its presence in the AER by this method. Thus, there is a slight possibility that CD44s detected by PCR in the AER-enriched tissue may have been amplified from a small amount of contaminating mesoderm or non-AER ectoderm. To rule out this possibility we performed Western blots of extracts obtained from AER-enriched tissue and the remaining limb tissues (mesodedectoderm). An -85-kDa pro- tein corresponding to CD44s was readily detected as the main CD44 isoform in the AER sample (Fig. 5). When equivalent samples of AER and mesodedectoderm were used, CD44s was seen to be much more concentrated in the AER sample than in the remaining tissues (compare lanes 1and 2 in Fig. 5).Thus, a t most, a small proportion of CD44s in the AER sample could have been derived from mesodermal or non-AER ectodermal contamination, and we conclude that CD44s is a major CD44 isoform in the AER. Distribution of Hyaluronan During Early Mouse Limb Development The results presented above indicate that one of the major CD44 isoforms expressed by the AER is CD44s, the hyaluronan receptor ( A d o et al., 1990; Culty et al., 1990).This finding suggests that the distribution of hyaluronan may be related in some way to the AER. Consequently we used the bPG probe (see Experimental Procedures) to determine whether we could detect hyaluronan in the mouse limb bud a t the time that the AER is present and expresses CD44s. We found staining for hyaluronan at a significant level throughout the mesoderm, but not in the the ectoderm, of E9 (Fig. 6B) and E10.5 (Fig. 6D,G) limb buds. At the latter stage there is a particularly strong reaction for hyaluronan in the basement membrane underlying the ectoderm of the limb buds. However this staining is interrupted beneath the AER (Fig. 6D,G). Strong reactivity for hyaluronan persists in the peripheral mesoderm of the limb bud (Figs. 611, but staining decreases markedly in the condensed mesoderm prior to cadilage differentiation (data not shown). We also compared the distribution of hyaluronan with CD44 by immunohistochemistry. As reported by Wheatley et al. (1993), immunoreactivity with antiCD44 mAb was strong within the AER of the E10.5 hindlimb (Fig. 6C,E,F) but was greatly reduced in the ectoderm of earlier (Fig. 6A) and later (Fig. 6H) stage limb buds; this distribution is in agreement with that of mRNA transcripts as obtained by in situ hybridization (Fig. 4). CD44 distribution as determined by immunohistochemistry or in situ hybridization was also concordant at later stages of limb development. Immunoreactivity at all stages was similar with or without hyaluronidase pretreatment of the sections. DISCUSSION Two outcomes of this study provide potentially new insights into AER function during limb bud morphogenesis. The first of these is identification of several CD44 isoforms that are highly enriched in the AER relative to mesoderm and other regions of the ectoderm of the early limb bud. The major CD44 isoforms detected in the AER, i.e., CD44s, V3-Vl0, V4-V10, and V6-Vl0, are likely to exhibit different biochemical activities and fulfill different functions in the AER. Our second striking observation is that the basement mem- CD44 ISOFORMS IN LIMB DEVELOPMENT Fig. 4. Localization of CD44 variant mRNAs to the AER by in situ hybridization. E l 0 mouse embryo limb buds were hybridized with digoxigenin-labeled riboprobes to common and variant regions of CD44. A: H3-5 antisense probe to wmmon region of CD44. B: V4-7 antisense probe. C: V4-5 antisense probe. In all three cases, as well as with the 209 V6-7, V6-10, and V8-10 antisense probes (not shown), strong positive signals were obtained in and restrictedto the AER. 0: H3-5 sense probe; no signal was detected in the AER or elsewhere in the limb. The same negative result was obtained with the other five sense probes (not shown). Bar = 40 Fm. several systems is mediation of hyaluronan uptake en route to degradation by lysosomal enzymes (Orkin et al., 1982; Culty et al., 1992, Hua et al., 1993; Underhill et al., 1993; Underhill, 1993; Pavasant et al., 1994). Possibly then, CD44s on the surface of AER cells mediates removal of hyaluronan from the basement membrane region directly beneath it. Consistent with this idea is the fact that lysosomal hyaluronidase is elevated in distal limb ectoderm relative to other regions of the limb (Kulyk and Kosher, 1987) and would thus degrade internalized hyaluronan in the AER region. Another hyaluronan-binding protein that has been implicated in endocytosis of hyaluronan is intercellular adhesion molecule-1 (ICAM) (McCourt et al., 1994). We have surveyed several stages of mouse embryo development for the distribution of ICAM and found that there is very low and diffuse expression in Fig. 5. identification of CD44s in extracts of AER and mesoderm. the limb bud but high levels in some other areas of the Extracts of AER-enriched tissue and the remaining limb mesoderm/ec- embryo (unpublished data). Thus, it is unlikely that toderm were analyzed by SDS-PAGE and Western blotting, using anti- ICAM would be responsible for removal of hyaluronan body KM201 against CD44. Lane 1:AER-enriched extract, 20 pg protein. Lane 2: Mesoderm/ectodermextract, 20 pg protein. Lane 3: Mesoderm/ from the sub-AER basement membrane. The depletion ectoderm extract, 50 pg protein. Lanes 4-6: Identical with lanes 1-3 in of hyaluronan in this region may facilitate AERmethe absence of primary antibody. The arrow indicates the position of soderm contact and subsequent morphogenesis in a migration of CD44s. similar manner to that of glycosaminoglycan turnover in the basement membrane during epithelial branching (Bernfield et al., 1984). Although it is clear that a major ligand for CD44s is brane region immediately beneath the AER is depleted of hyaluronan relative to the basement membrane un- hyaluronan, this is not the case for the variant CD44 isoforms (Jackson et al., 1995). We have found that the derlying other regions of the ectoderm. One of the most prominent isoforms of CD44 in the same CD44 isoforms as described in this study are also AER is CD44s, a well-characterized hyaluronan recep- present in many other inductive epithelia undergoing tor. A role that has been ascribed to this receptor in active morphogenesis, including epithelia of develop- 210 W ET AL. Fig. 6. Distributionof hyaluronan and CD44 proteinduring early mouse limb development. In A,C,E,F, and H, mAb KM201 was used for detection of CD44. In B,D,G,I, hyaluronan was stained with bPG. A: Longitudinal section of E9 forelimb. Arrows indicate dispersed CD44-positive cells in the mesoderm. The rest of the mesoderm and the ectoderm are virtually negative. 6: Longitudinal section of E9 forelimb showing hyaluronan present throughout the mesoderm but absent from the ectoderm (arrowheads). C,D: Cross sections of E10.5 hindlimb at high power, showing the presenceof CD44 and the absenceof hyaluronan,respectively, in the AER (arrowheads). Hyaluronan staining is strong in the basement membrane (solid arrows) underlying the ectoderm except beneath the AER (open arrows), where it is absent. E,F: Longitudinaland cross sections, respectively, of E10.5 hindlimb showing that CD44 is restricted to the AER (arrowheads)and a few dispersed cells in the mesoderm. G: Cross section of E10.5 hindlimb showing that hyaluronan is distributed throughout the mesoderm, is highly enriched in the basement membrane (solid arrows) except beneath the AER (open arrows), and is absent from the entire ectoderm including the AER (arrowheads). H,I: Longitudinal sections of E l 2 hindlimb, showing the absence of CD44 immunoreactivity from the ectoderm and mesoderm and the presence of hyaluronan in the mesoderm but not the ectoderm. Bar in I = 80 pm and applies to A,B,F,G and H: bar in D = 25 pm and applies to C; bar in E = 100 pm. ing teeth, nose, ear, and hair follicles (Yu and Toole, unpublished data). Some of these CD44 variants are also expressed by adult epithelia (Terpe et al., 1994; Mackay et al., 19941, and many are expressed by malignant carcinoma cells where they are thought to confer metastatic properties to these cells (Gunthert et al., 1991; Rudy et al., 1993). The ligands and cellular roles of most of these variants are not established, but it seems very likely that they are important in some aspect of morphogenetic cell behavior. One variant of particular interest is CD44 (V3-V10), which was found to be present at readily detectable levels in AER-enriched tissue. The V3 exon encodes the optimal consensus motif for addition of heparan sulfate side chains, and thus V3-containing isoforms are usually expressed as heparan sulfate proteoglycans. These proteoglycans bind FGF avidly (Jackson et al., 1995; Bennett et al., 1995). Since members of the FGF family are localized within the AER and are critical for its morphogenetic action (Niswander et al., 1993; Fallon et al., 1994; Crossley and Martin, 1995; Ochiya et al., 1995), heparan sulfate-conjugated CD44 (V3-VlO) may be involved in retaining FGF preferentially at the surface of AER cells and presenting it in an appropriate configuration to promote cellular proliferation in the mesoderm immediately underlying the AER or within the AER itself (Rapraeger et al., 1991; Yayon et al., 1991; Spivak-Kroizman et al., 1994). It is interesting to note that other heparan sulfate-proteoglycans that might participate in such events, i.e., syndecan and perlecan, are not present in the AER (Solursh and Jensen, 1988; Solursh et al., 1990; Gould et al., 1992). Even though FGF-2 is produced throughout most of the early limb bud ectoderm (Savage et al., 19931, it may be unable to stimulate proliferation outside the AER, or the region of mesoderm immediately subjacent to the CD44 ISOFORMS IN LIMB DEVELOPMENT AER, because of the absence of CD44 (V3-VlO) from non-AER limb ectodenn, The latter point together with restriction of other fibroblast growth factors to the AER (Niswander et al., 1993; Crossley and Martin, 1995)could provide part of the mechanism whereby the AER controls directional outgrowth, as has been established in previous biological experiments (Saunders, 1948; Summerbell et al., 1973). Although it is still not clear how FGF-2 is secreted, the function of FGF-2 in non-AER ectoderm may be to regulate homeobox gene expression, e.g., Msx-1 (Wang and Sassoon, 1995), or extracellular matrix composition in the subectodermal mesoderm (Knudson et al., 1995) so as to facilitate the differentiation program unique to this region. The role of other CD44 variants in the AER, especially V4-VIO and V6-VIO which are present in considerable amounts, is not clear. However, their restricted localization to the AER and to other instructive epithelia (Yu and Toole, unpublished data) suggests strongly that they participate in the function or maintenance of these specialized epithelia, possibly through interactions with other morphogenetic factors, extracellular matrix macromolecules, or cell adhesion components. 211 TABLE 2. PCR Primers for In Situ Probesa Snnw ------ H3-5: 5'-CGCCATGGACUGTMTGGT-3' V4-5: 5'-TTGCAAGTACTCCACGGGTT-3' V4-7: 5'-TTGCAAGTACTCCACGGG'IT-3' V6-7: 5'-CTCCTAATAGTACAGCAGAA-3' V6-1 0:5'-GTGAAGACTCCCATGTGACA-3' V8-10: 5'-ATACAGACTCCAGTCATAGT-3' Antisense H3-5: 5'-GTGACTGATGTACAGTC'N'C-3' V4-5: 5'-TTGTGC'ITGTAGCATGTGGG3' V4-7: 5'-TGCITTCTG'r"ITGATGACC-3' V6-7: 5'-TGCTTTCTG'ITTGATGACC-3' V6-10: 5'-AGTTGTATCTGTGGGUGAC-3' V8-10: 5'-CTGGTAAGGAGCCATCAACA-3' The sequences were derived from Screaton et al. (1993) (GenbanmMBL accession number L13611). cDNA Cloning and Sequencing The products from PCR were either directly cloned into the pCRII cloning vector (Version 2.3, Invitrogen, San Diego, CA) or purified using Geneclean (BiolOl Inc, La Jolla, CA) from 1%agarose gels, then cloned into the pCRII cloning vector. The nucleotide sequences of the inserts were determined using the dideoxy chain termination method with Sequenase (US EXPERIMENTAL PROCEDURES Biochemical Corp., Cleveland, OH) from doubleDetection of CD44 Isoforms by RT-PCR stranded templates following the manufacturer's instructions. Total RNA was isolated from mouse embryonic tissues using TRIzol reagent (Gibco-BRL)according to the Preparation of Tissue Sections manufacturer's instructions. Embryos were obtained from pregnant Swiss WebcDNA was synthesized using Superscript I1 RNase H- reverse transcriptase (Gibco-BRL) following the ster mice (Taconic Laboratory, Germantown, NY), and manufacturer's instructions. Briefly, 5 pg of total RNA embryonic age was determined from the time of apwas first incubated with 1 pl of 500 p.g/ml oligo pearance of a vaginal plug. Embryos of different stages (dT),z-le (Gibco-BRL)at 70°C for 10 minutes, then with were fixed in 4% paraformaldehyde (Tousimis, RockSuperscript I1 RNase H- reverse transcriptase at 37°C ville, MD) in phosphate-buffered saline (PBS), washed for 1hour, followed by 1p1 of RNase H (2.7 U/p1; Gibco- in PBS, dehydrated through 30%,70%, 95%, 100%ethBRL) at 37°C for 15 minutes to degrade the RNA tem- anol and xylene, and then embedded in paraffin wax (Fisher, Columbia, MD). Sections (6-8 pm) were cut, plate. mounted on slides, and used for immunohistochemisTwo 20-mer oligonucleotides, primers 1and 2, corresponding to sequences that flank the variant exons of try, hyaluronan-affinity histochemistry, or in situ hyCD44 (see Table l), were used for the first set of PCR bridization. experiments. In the second set of PCR experiments, sense primer 2 was used with antisense primers from In Situ Hybridization Six different sets of primers, the sequences of which exons v10, v9, v8, v7, v6, v5, v4, v3. In the third set of PCR experiments, antisense primer 1 was used with are given in Table 2, were used in PCR to generate six sense primers from exons v l , v2, v3, v4, v5, v6, v7, v8, different probes. These probes are 1)H3-5, composed v9, v10 (see Table 1 for details). The variant exon se- of nucleotides 6-416 of CD448, which is common to all quences were derived from Screaton et al. (1993) (Gen- isoforms of CD44; 2) V4-5, spanning exons V4 through bankEMBL accession number L13611). PCR was per- V5; 3) V4-7, spanning exons V4 through V7; 4) V6-7, formed using Ampli Taq DNA polymerase (Perkin spanning exons V6 through V7; 5) V6-10, spanning Elmer, Norwalk, CT) according to the manufacturer's from within exon 6 to within exon 10; and 6) V8-10, instructions. After 35 cycles a t 94°C for 30 seconds, spanning exons V8 through V10. The PCR products 55°C for 30 seconds, and 72°C for 1minute, followed by were cloned into pCRII vector separately, and then a a final 5 minutes a t 72"C, 10 p,1 of the PCR reaction single clone was picked for each insert and sequenced products were analyzed by 1% agarose gel electro- to confirm authenticity and orientation. Plasmids containing the inserts were prepared using phoresis. The remaining portions of each sample were a Qiagen plasmid Midikit (Qiagen Inc, Chatsworth, used for cDNA cloning. 2 12 YU ET AL. CA). These plasmids were linearized using BamHI or XhoI (Gibco-BRL)and transcribed with T7 or SP6 RNA polymerase (Promega, Madison, WI) in the presence of digoxigenin-conjugated UTP (Boehringer-Mannheim, Indianapolis, IN) to make sense or anti-sense digoxigenin riboprobes. After transcription a t 37°C for 2 hours, RQ1 DNase (Promega) was used to degrade the DNA template. Digoxigenin-riboprobes were precipitated in the presence of LiCl and ethanol. The quantity of labeled RNA was checked by dot-blotting with alkaline phosphatase-conjugated polyclonal antibody to digoxigenin (Boehringer-Mannheim) according to the manufacturer’s instructions. Before hybridization, the probes were denatured by heating a t 65°C for 10 minutes. In situ hybridization was performed by adapting procedures from Rider et al. (19921, Wilkinson (1992j, and the manufacturer’s protocols (Boehringer-Mannheim). Briefly, paraffin sections were deparaffinized and rehydrated, treated with proteinase K and refixed with formaldehyde, acetylated with acetic anhydride, and dehydrated before hybridization. Hybridization was performed overnight at 60°C. Nonspecifically bound probe was digested with RNase, and sections were washed in 50% formamide/lx SSC a t 60°C. Positive signals were detected using anti-digoxigenin antibody conjugated with alkaline phosphatase and chromogenic substrates NBT and BCIP according to the manufacturer’s instructions (Boehringer-Mannheim). The sections were mounted in gel mount (Biomedia Corp, Foster City, CA). Hyaluronan-AffinityHistochemistry The hyaluronan-binding region of cartilage proteoglycan (aggrecan) binds strongly and specifically to hyaluronan, and thus can be used as a probe for detection of hyaluronan in cells and tissue sections (Knudson and Toole, 1985; Green et al., 1988).Thus, we used the biotinylated hyaluronan-binding region of cartilage proteoglycan (bPG, kindly provided by Dr. Charles B. Underhill) for this purpose. After inactivation of endogenous peroxidase as for immunohistochemistry , sections were incubated with 2 pg/ml bPG in PBS containing 10% calf serum (Gibco-BRL, Gaithersburg, MD) for 30 minutes a t room temperature. To determine background staining, 2 pg/ml bPG was first mixed with 100 pg/ml hyaluronan before use; virtually no staining was obtained when this control was used. After extensive washing with PBS, bound bPG was detected using Vector A and B reagents (Vector Laboratories) and diaminobenzidine tetrahydrochloride dihydrate (Aldrich) according to the manufacturers’ instructions. Tissue Extraction and Western Blotting Approximately 200 limbs from E l 0 mouse embryos were dissected into two parts: AER-enriched tissue and the remaining limb tissues (mesoderdectoderm) (see Fig. 1A). The two tissue groups were homogenized in 5 volumes of 5 mM HEPES, 2 mM MgCI,, pH 7.4, containing 2 mM PMSF, 2 pg/ml leupeptin, and 0.5 U/ml aprotinin in Dounce homogenizers. Nuclei and cellular debris were removed by centrifugation at 6,300g for 5 minutes. The supernatant was centrifuged at 14,OOOg for 10 minutes, and the membrane pellets were lysed in 50 mM Tris-HC1 buffer, pH 7.4, containing 150 mM NaCl, 5 mM EDTA, 1%Triton X-100, 0.1% sodium dodecyl sulfate (SDS),2 mM phenylmethylsulfonyl fluoride (PMSF), 2 pg/ml leupeptin, and 0.5 U/ml aprotinin. The membrane extracts were subjected to 10% SDSpolyacrylamide gel electrophoresis under nonreducing conditions, then transferred to Hybond-ECL nitrocellulose membranes (Amersham, Arlington Heights, IL). After transfer, the membranes were blocked in 1% BSA, 1%nonfat milk, and 5 pg/ml goat IgG in 0.2% Tween 20,0.05 M Tris-HCL, 0.15 M NaC1, pH 8.0, then incubated with monoclonal antibody, KM201, against CD44 (ATCC, Rockville, MD) a t 4°C overnight. After extensive washing with 0.2% Tween 20, 0.05 M TrisHCl, 0.15 M NaC1, pH 8.0, the membranes were incubated with goat anti-rat secondary antibody conjugated with horseradish peroxidase for 30 minutes a t room temperature, followed by ECL detection reagents (Amersham) according to the manufacturer’s instructions. CD44 Immunohistochemistry For localization of CD44, we used monoclonal antibody (mAb) KM201 prepared from hybridoma cultures (ATCC, Rockville, MD) or mAb 5D2-27 obtained form Developmental Studies Hybridoma Bank (NICHD, Baltimore, MD). Partially purified KM201, obtained by 50% (NH,),SO, precipitation of KM201 hybridoma culture medium was used in the experiments reported here unless mentioned otherwise. After the sections were rehydrated through a series of ethanol concentrations and water, they were incubated in 3% H,O, in methanol for 1hour at room temperature to inactivate endogenous peroxidase, and with 1%bovine serum albumin (BSA), l%nonfat milk, 10 pg/ml goat-IgG at room temperature for 1 hour to block non-specific staining. The sections were then incubated with 5-10 Fg/ml KM201 a t 4°C overnight. Before treatment of 3% H202, some sections were treated with testicular hyaluronidase (50 units/ml in 0.15 M NaC1, 0.05 M Na acetate, pH 5.0) a t 37°C for 3 hours, but little difference in staining was observed. Biotinylated goat anti-rat secondary antibody followed by Vector staining reACKNOWLEDGMENTS agents A and B (Vector Laboratories, Burlingame, CA) We thank Drs. Tet-kin Yeo and Stuart Tobet for critand diaminobenzidine tetrahydrochloride dihydrate (Aldrich Chemical Company, Milwaukee, WI) were ical reading of the manuscript. This research was supused according to the manufacturers’ instructions to ported by grants DE 05838 and HD 23681 from the detect KM201 reactivity. NIH (B.P.T.). CD44 ISOFORMS IN LIMB DEVELOPMENT REFERENCES Arch, R., Wirth, K., Hofmann, M., Ponta, H., Matzku, S., Herrlich, P., and Zoller, M. (1992) Participation in normal immune responses of a metastasis-inducing splice variant of CD44. Science 257682485, Aruffo, A., Stamenkovic, I., Melnick, M., Underhill, C.B., and Seed, B. (1990) CD44 is the principal cell surface receptor for hyaluronate. Cell. 61:1303-1313. Bennett, K.L., Jackson, D.G., Simon, J.C., Tanczos, E., Peach, R., Modrell, B., Stamenkovic, I., Plowman, G., and Aruffo, A. (1995) CD44 isoforms containing exon V3 are responsible for the presentation of heparin-binding growth factor. J. Cell Biol. 128:687698. Bernfield, M., Banerjee, S.D., Koda, J.E., and Rapraeger, A.C. (1984) Remodeling of the basement membrane as a mechanism of morphogenetic tissue interaction. In: “The Role Extracellular Matrix in Development,” Trelstad R.L. (ed). New York Alan Liss, Inc., pp 545-572. Bourguignon, L.Y.W., Lokeshwar, V.B., Chen, X., and Kerrick, W.G.L. (1993) Hyaluronic acid-induced lymphocyte signal transduction and HA receptor (gp85/CD44)-cytoskeletoninteraction. J. Immunol. 151:6634-6644. Crossley, P.H., and Martin, G.R. (1995) The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121:439-451. Culty, M., Miyake, K., Kincade, P.W., Sikorski, E., Butcher, E.C., and Underhill, C.B. (1990) The hyaluronate receptor is a member of the CD44 (H-CAM) family of cell surface glycoproteins. J. Cell Biol. 111:2765-2774. Culty, M., Nguyen, H.A., and Underhill, C.B. (1992) The hyaluronan receptor (CD44) participates in the uptake and degradation of hyaluronan. J . Cell Biol. 116:1055-1062. Faassen, A.E., Schrager, J.A., Klein, D.J., Oegema, T.R., Couchman, J.R., and McCarthy, J.B. (1992) A cell surface chondroitin sulfate proteoglycan, immunologically related to CD44, is involved in type I collagen-mediated melanoma cell motility and invasion. J. Cell Biol. 116:521-531. Fallon, J.F., Lopez, A., Ros, M.A., Savage, M.P., Olwin, B.B., and Simandl, B.K. (1994) FGF-2 apical ectodermal ridge growth signal for chick limb development. Science 264104-107. Fenderson, B.A., Stamenkovic, I., and Aruffo, A. (1993) Localization of hyaluronan in mouse embryos during implantation, gastrulation and organogenesis. Differentiation 54:85-98. Goldstein, L.A., Zhou, D., Picker, L., Minty, C.N., Bargatze, R.F., Ding, J.F., and Butcher, E.C. (1989) A human lymphocyte homing receptor, the Hermes antigen, is related to cartilage proteoglycan core and link proteins. Cell 561063-1072. a u l d , S.E.,Upholt, W.B., and Kosher, R.A. (1992) Syndecan 3: a member of the syndecan family of membrane-intercalated proteoglycans that is expressed in high amounts at the onset of chicken limb cartilage differentiation. Roc. Natl. Acad. Sci. U.S.A. 89: 3271-3275. Green, S.J., Tarone, G., and Underhill, C.B. (1988) Distribution of hyaluronate and hyaluronate receptors in the adult lung. J. Cell Sci. 89145-156. Gunthert, U., Hofmann, M., Rudy, W., Reber, S., Zoller, M., Haussmann, I., Matzku, S., Wenzel, A., Ponta, H., and Herrlich, P. (1991) A new variant of glycoprotein CD44 confers metastatic potential to rat carcinoma cells. Cell 6.513-24. He, Q., Lesley, J., Hyman, R., Ishihara, K., and Kincade, P.W. (1992). Molecular isoforms of rnurine CD44 and evidence that the membrane proximal domain is not critical for hyaluronate recognition. J . Cell Biol. 1191711-1719. Hua, Q., Knudson, C.B., and Knudson, W. (1993) Internalization of hyaluronan by chondrocytes occurs via receptor-mediated endocytosis. J . Cell Sci. 106:365-375. Jackson D.G., Bell, J., Dickinson, R., Timans, J., and Shields, J. (1995) Proteoglycan forms of the lymphocyte homing receptor CD44 are alternatively spliced variants containing the V3 exon. J. Cell Biol. 128:673-685. Jalkanen, S., and Jalkanen, M. (1992) Lymphocyte CD44 binds the 213 COOH-terminal heparin-binding domain of fibronectin. J. Cell Biol. 116:817-825. Kennel, S.J., Lankford, T.K., Foote, L.J., Shinpock, S.G., and Stringer, C. (1993) CD44 expression on murine tissues. J Cell Sci. 104:373-382. Knudson, C.B., and Toole, B.P. (1985)Fluorescent morphological for hyaluronate. J. Cell Biol. 1001753-1758. Knudson, C.B., Munaim, $.I., and Toole, B.P. (1995) Ectodermal stimulation of the production of hyaluronan-dependent pericellular matrix by embryonic limb mesodermal cells. Dev. Dyn. 204186-191. Kulyk, W.M., and Kosher, R.A. (1987) Temporal and spatial analysis of hyaluronidase activity during development of the embryonic chick limb bud. Dev. Biol. 120:535-541. Lacy, B.E., and Underhill, C.B. (1987) The hyaluronate receptor is associated with actin filaments. J. Cell Biol. 105:1395-1404. Lesley, J., He, Q., Miyake, K., Hamann, A., Hyman, R., and Kincade, P.W. (1992) Requirements for hyaluronic acid binding by CD44: a role for the cytoplasmic domain and activation by antibody. J. Exp. Med. 175257-266. Lesley, J., Hyman, R., and Kincade, P. (1993) CD44 and its interaction with extracellular matrix. Adv. Immunol. 54371-335. Mackay, C.R., Terpe, H.J., Stauder, R., Marston, W.L., Stark, H., and Gunthert, U. (1994) Expression and modulation of CD44 variant isoforms in humans. J Cell Biol. 124:71-82. McCourt, P.A., Ek, B., Forsberg, N., and Gustafson, S. (1994) Intercellular adhesion molecule-1 is a cell surface receptor for hyaluronan. J. Biol. Chem. 269:30081-30084. Neame, S.J., and Isacke, C.M. (1992) Phosphorylation of CD44 in vivo requires both Ser323 and Ser325, but does not regulate membrane localization or cytoskeletal interaction in epithelial cells. EMBO J. 11:4733-4738. Niswander, L., Tickle, C., Vogel, A., Booth, I., andMartin, G.R. (1993) FGF-4 replaces the apical ectodermal ridge and directs outgrowth and patterning of the limb. Cell 75:579-587. Ochiya, T., Sakamoto, H., Tsukamoto, M., Sugimura, T., and Terada, M. (1995) Hst-1 (FGF-4) antisense oligonucleotides block murine limb development. J . Cell Biol. 130:997-1003. Orkin, R.W., Underhill, C.B., and Toole, B.P. (1982) Hyaluronate degradation in 3T3 and simian virus-transformed 3T3 cells. J. Biol. Chem. 257:5821-5826. Pavasant, P., Shizari, T.M., and Underhill, C.B. (1994) Distribution of hyaluronan in the epiphysial growth plate: turnover by CD44-expressing osteoprogenitor cells. J. Cell Sci. 107:2669-2677. Fkpraeger, A.C., Krufka, A., and Olwin, B.B. (1991) Requirement of heparan sulfate for bFGF-mediated fibroblast growth and myoblast differentiation. Science 252:1705-1708. Rider, V., Carlone, D.L., Cai, C., and Oliver, N. (1992) Uterine fibronectin mRNA content and localization are modulated during implantation. Dev. Dyn. 195:l-14. Rudy, W., Hofmann, M., Schwartz-Albiez, R., Zoller, M., Heider, K.H., Ponta, H., and Herrhch, P. (1993)The two major CD44 proteins expressed on a metastatic rat tumor cell line are derived from different splice variants: each one individually suffices to confer metastatic behavior. Cancer Res. 53:1262-1268. Ruiz, P., Schwarzler, C., and Gunthert, U.(1995) CD44 isoforms during differentiation and development. BioEssays 17:17-24. Saunders, J.W. (1948) The proximo-distal sequence of origin of the parts of the chick wing and the role of the ectoderm. J. Exp. Zool. 108:363-403. Savage, M.P., Hart, C.E., Riley, B.B., Sasse, J., Olwin, B.B., and Fallon, J.F. (1993) Distribution of FGF-2 suggests it has a role in chick limb bud growth. Dev. Dyn. 198159-170. Screaton, G.R., Bell, M.V., Bell, J.I., and Jackson, D.G. (1993) The identification of a new alternative exon with highly restricted tissue expression in transcripts encoding the mouse Pgp-1 (CD44) homing receptor. J. Biol. Chem. 268:12235-12238. Sherman, L., Sleeman, J., Herrlich, P., and Ponta, H. (1994) Hyaluronate receptors: key players in growth, differentiation, migration and tumor progression. Curr. Opin. Cell Biol. 6:726-733. Solursh, M., and Jensen, K. (1988) The accumulation of basement membrane components during the onset of chondrogenesis and myogenesis in the chick wing bud. Development 104:41-49. 214 YU ET AL Solursh, M., Reiter, R.S., Jensen, K.L., Kato, M., and Bernfield, M. (1990) Transient expression of a cell surface heparan sulfate proteoglycan (syndecan) during limb development. Dev. Biol. 140:8392. Spivak-Kroizman, T., Lemmon, M.A., Dikic, I., Ladbury, J.E., Pinchasi, D., Huang, J.,Jaye, M., Crumley, G., Schlessinger, J., and Lax, I. (1994) Heparin-induced oligomerization of FGF molecules is responsible for FGF receptor dimerization, activation, and cell proliferation. Cell 791015-1024. Stamenkovic, I., Amiot, M., Pesando, J.M., and Seed, B. (1989) A lymphocyte molecule implicated in lymph node homing is a member of the cartilage link protein family. Cell 56:1057-1062. Summerbell, D., Lewis, J.H., and Wolpert, L. (1973) Positional information in chick limb morphogenesis. Nature 244:492-496. Tanaka, Y., Adams, D.H., Hubscher, S., Hirano, H., Siebenlist, U., and Shaw, S. (1993)T-cell adhesion induced by proteoglycan-immobilized cytokine MIF-lp. Nature 361:79-82. Terpe, H.-J., Stark, H., Prehm, P., and Gunthert, U. (1994) CD44 variant isoforms are preferentially expressed in basal epithelia of non-malignant human fetal and adult tissues. Histochemistry 101: 79-89. Toole, B.P., Banerjee, S.D., Turner, R., Munaim, S., and Knudson, C. (1991) Hyaluronan-cell interactions in limb development. In: “Developmental Patterning of the Vertebrate Limb,” Hinchliffe, J.R., Hurle, J.M., and Summerbell, D. (eds). New York Plenum Press, pp 215-223. Tsukita, S., Oishi, K., Sato, N., Sagara, J., Kawai, A., and Tsukita, S. (1994) ERM family members as molecular linkers between the cell surface glycoprotein CD44 and actin-based cytoskeletons. J. Cell Biol. 126:391-401. Underhill, C.B. (1993) Hyaluronan is inversely correlated with the expression of CD44 in the dermal condensation of the embryonic hair follicle. J. Invest. Dermatol. 101:820-826. Underhill, C.B., Nguyen, H.A., Shizari, M., and Culty, M. (1993) CD44 positive macrophages take up hyaluronan during lung development. Dev. Biol. 155:324-336. Wang, Y., and Sassoon, D. (1995) Ectoderm-mesenchyme and mesenchyme-mesenchyme interactions regulate Msx-1 expression and cellular differentiation in the murine limb bud. Dev. Biol. 168374382. W’heatley, S.C., Isacke, C.M., and Crossley, P.H. (1993) Restricted expression of the hyaluronan receptor, CD44, during postimplantation mouse embryogenesis suggests key roles in tissue formation and patterning. Development 119295-306. Wilkinson, D.G. (1992) “In Situ Hybridization: A Practical Approach.” New York: IF& Press. Yayon, A., Klagsbrun, M., Esko, J.D., Leder, P., and Ornitz, D.M. (1991) Cell surface, heparin-like molecules are required for binding of basic fibroblast growth factor to its high affinity receptor. Cell 64:841-848. Zhou, D.F., Ding, J.F., Picker, L.J., Bargatze, R.F., Butcher, E.C., and Goeddel, D.V. (1989) Molecular cloning and expression of Pgpl: the mouse homolog of the human HCAM (Hermes) lymphocyte homing receptor. J. Immunol. 143:3390-3395.