In situ hybridization (ISH) and immunocytochemistry were used to localize sites of synthesis and deposition of the basement membrane glycoprotein laminin during development in the postimplantation mouse embryo and extraembryonic membranes. In addition, similar studies were performed on postnatal viscera during the first 20 days after birth. Up to 10 days post coitum, embryonic laminin synthesis was confined to parietal endoderm. In maternal tissue, intense laminin mRNA expression was detected in decidual cells in the mesometrial and antimesometrial endometrium at 5–7 days. At 10 days, uniform expression was still seen within the mesometrial endometrium, with higher levels around migrating trophoblast, but in the antimesometrial aspect expression was restricted to the basal zone. High levels of mRNA expression persisted in parietal endoderm throughout gestation but much lower levels were detected in visceral yolk sac. In the mature placenta, laminin mRNA expression was also found associated with fetal vessels in the labyrinth and giant cells at the fetal/maternal boundary. In the embryo, the external limiting membrane of the cerebral vesicles and spinal cord stained for laminin protein and detectable mRNA was found in the pia mater. Growing peripheral nerves and dorsal and ventral root fibres expressed laminin mRNA and stained for laminin protein. Laminin mRNA expression was found in ureteric buds and nephrogenic vesicles (but not in metanephric blastema) during early prenatal kidney development, and in glomeruli, Bowman’s capsule, loops of Henle and collecting duct cells at later stages of development, and after birth. All these structures possessed laminin-rich basement membrane (BM). Laminin mRNA expression fell to below detectable levels in the kidney around weaning. In the gut, laminin expression and protein staining was confined to the muscularis externa and the lamina propria during embryogenesis. After birth, the muscularis externa, muscularis mucosa and lamina propria cells corresponding to fibroblasts had detectable laminin mRNA, but in adult gut no laminin mRNA could be demonstrated in any cell type. In liver, low levels of laminin mRNA were seen in the capsule and in periportal connective tissue. After birth, laminin mRNA was associated with intrahepatic bile channels; no laminin mRNA was detected in the parenchyma and protein deposition was restricted to blood sinus BM. In the adult liver, no laminin mRNA was detected in any cell type. The developing heart showed uniform expression of laminin mRNA from 12 days to before birth. Postnatally, labelling was restricted to connective tissue cells.

Laminin is a glycoprotein found exclusively in basement membrane (BM) and was originally isolated from the Engelbreth-Holm-Swarm (EHS) tumour (Timpl et al. 1979). EHS cells synthesize four polypeptides A, B1a, B1b and B2 which are assembled via disulphide bonds into the mature laminin protein (Mr∼900000). Rotary shadowing of laminin reveals an asymmetric cross structure with globular domains at the tip of each arm (Engel et al. 1981; Rao et al. 1982). There is evidence that the protein secreted by non-neoplastic cells contains an extra globular domain at the end of the long arm corresponding to a fifth polypeptide designated M (Ohno et al. 1985), and that other isoforms of the protein may also exist (reviewed by Martin & Timpl, 1987). Laminin has binding domains for heparin (Ott et al. 1982) and type IV collagen (Rao et al. 1982), and it also possesses a domain which promotes neurite outgrowth in vitro ; this is probably located near the end of the long arm close to the heparin-binding domain (Davis et al. 1985).

Cells are able to bind to laminin via a specific 67 × 103Mr cell receptor (Rao et al. 1983), which recognizes a domain located in the central region of the cross structure. Recently, a stretch of nine amino acids in the Bl polypeptide chain which interacts with this receptor has been identified (Graf et al. 1987). A second cell-binding domain has been described near the end of the long arm of the cross, which appears to be able to interact with the 140 × 103Mr avian fibronectin receptor (integrin, CSAT antigen; Horwitz et al. 1985). Mouse parietal endoderm cells, derived from teratocarcinoma cells cultured in vitro, express a similar 140 × 103Mr receptor whose interaction with both fibronectin and laminin is disrupted by a synthetic peptide RGDS, which is known to block fibronectin binding to the CSAT antigen (Horwitz et al. 1985), and by antibodies to integrin (Grabel & Watts, 1987). Despite the observed interaction of laminin with integrin-like receptors (reviewed by Buck & Horwitz, 1987), the RGDS peptide sequence involved in receptor binding has not yet been located in the protein structure of laminin (Martin & Timpl, 1987). Trophoblast cells will also adhere to laminin, but, although they express a 140× 103Mr fibronectin receptor, their attachment to laminin is not blocked by peptides containing the RGDS sequence (Armant et al. 1986b), and a 140× 103Mr fibronectin receptor has been isolated from mammalian cells which does not bind laminin (Pytela et al. 1985). It remains to be determined whether the adhesion of trophoblast to laminin is mediated via the 67× 103Mr receptor or another receptor specific for laminin. Thus, although it is clear that mouse embryonic cells interact in vitro with laminin, the precise receptors involved have not been fully elucidated.

In the mouse, laminin is the first demonstrable extracellular protein synthesized during embryogenesis, being detectable on the cell surface by immunofluorescence after the first postfertilization division (Dziadek & Timpl, 1985). As development proceeds, laminin is detectable in the earliest stages of BM formation. Studies on the morphogenesis of kidney tubules have shown that laminin is present in differentiating renal tissue, e.g. ureteric buds and metanephric vesicles, but not in undifferentiated mesenchyme (Ekblom et al. 1980; Sariola & Ekblom, 1985). Laminin is thought to have an important role in neuronal development and regeneration (reviewed by McClay & Ettensohn, 1987; Martin & Timpl, 1987), and studies in vitro have demonstrated nerve outgrowth from both peripheral and central nervous system (CNS) neurones on laminin-containing matrices (Rogers et al. 1983). Extracts of embryonic muscle support the survival and outgrowth of motor neurones in culture, and the proteins responsible for this activity are immunoprecipitated by laminin antibodies (Dohrmann et al. 1986). Laminin accumulation in the region of regenerating neurones has been observed in vivo (Longo et al. 1984). Avian neurites express the CSAT antigen and an antibody to this receptor protein blocks chick neurite outgrowth on laminin-containing matrices in vitro (Bozyczko & Horwitz, 1986; Cohen et al. 1987). Laminin can also mediate transdifferentiation of retinal pigmented epithelial cells into neurones in vitro (Reh et al. 1987). Results like these support the theory that laminin is the major component of Neurite Promoting Factors (NPFs; Davis et al. 1985), and has a major role in neuronal growth and differentiation. In contrast, Coughlin and co-workers (Coughlin et al. 1986) isolated the major functional NPF from mouse heart cell cultures and found it to be a smaller (350 ×103Mr) molecule which they termed neuronectin. As yet the immunological and structural relationship between neuronectin and laminin is not known.

In the rat embryo, laminin immunoreactivity has been found in the endoderm BM of the extraembryonic membranes (Tuckett & Morriss-Kay, 1986), and also in trophoblastic and vascular BM in human placenta (Kurosawa et al. 1985; Amenta et al. 1986).

In this communication, we report the use of in situ hybridization (ISH) to localize cells expressing laminin mRNA in the postimplantation embryo, the placenta and in early postnatal life of the mouse, with the object of identifying which cell types are responsible for laminin production. In addition, we have used a well-characterized polyclonal antiserum to the laminin protein (Forster et al. 1984), and a sensitive indirect immunoperoxidase technique to relate sites of protein deposition to cells expressing mRNA.

Probe preparation

A 675 bp EcoRI/SalI laminin cDNA, derived from a parietal endoderm cDNA library (Barlow et al. 1984) and encoding part of the B2 polypeptide chain, was cloned into the Bluescript (Stratagene) SK plasmid. This vector has a T7 and a T3 RNA polymerase promoter sequence, one located at each end of the polylinker, permitting either strand of the cloned DNA to be transcribed from the same plasmid. Bluescript SK containing the laminin cDNA was used to transform E. coli strain XLB-1 (Stratagene) and pure plasmid was prepared by standard CsCl gradient methods (Maniatis et al. 1982). The plasmid was linearized by digestion with the appropriate restriction endonuclease at a site in the polylinker sequence downstream of the insert with respect to the RNA polymerase promoter to be used. High-specific-activity transcript was produced by the following method; at room temperature, 5 μ l 5 × Transcription buffer (Stratagene), 2 μ l 100mm-DTT, I μ l each 10mm-ATP, -CTP, -GTP, 0 ·5mm-UTP and 12 μ l (120 μ Ci) 35S-UTP (∼ 1000 Ci mmol-1, Amersham) were mixed in a 0 ·5 ml centrifuge tube. 1 μ l of l μ g μ l-1 linearized template was added followed by 1 μ l of the appropriate RNA polymerase, either T3 (Stratagene) or T7 (Boehringer). The reaction was allowed to proceed for 40 min at 37 °C. A further l μ l of polymerase was added and the incubation continued for a further 40 min. 2 μl of 20 mg ml-1 yeast RNA (protease digested, phenol extracted), 2 μ l human placental ribonuclease inhibitor, and 0 · 5 μ l of RNase-free DNase I were then added. After a further 10 min incubation at 37°C, 1 μ l of 1 M-DTT was added, the volume made up to 108 μ l and samples removed for trichloroacetic acid (TCA) precipitation and electrophoresis. The probe was precipitated in absolute ethanol at – 70°C for at least 2h, then centrifuged and the pellet dried. The probe was resuspended in 100 μ l of 10 mm-DTT and then hydrolysed to an average length of 150 – 300 nucleotides (Cox et al. 1984), ethanol precipitated, resuspended at 5ng μ l-1 in 10 mm-DTT, and stored at – 70 °C.

Following TCA precipitation and liquid scintillation counting of the sample (Maniatis et al. 1982), the yield of probe was calculated. Transcript quality was assessed by electrophoresis through 3·5 % polyacrylamide gels containing 20% formamide.

Tissue preparation

Embryos and placentae were fixed in 4% paraformaldehyde in Dulbecco’s phosphate-buffered saline (PBS; without Ca2+ or Mg2+) for 18 – 24h at room temperature. The age of the embryos was estimated from the first appearance of the copulation plug (day 1). 5-to 10-day embryos were fixed in utero to aid orientation. Embryos from 11 to 19 days were dissected from the uterus and extraembryonic membranes prior to fixation. Fixative was injected into the peritoneal cavity of embryos older than 15 days post coitum. Mouse pups from immediately after birth to 22 days postnatal and adult mice were killed by cervical dislocation following pentrane anaesthesia. The visceral and thoracic organs were dissected out and fixed as above. Following fixation, the tissues were transferred to 0 · 5 m-sucrose in PBS and stored at 4 °C prior to dehydration and wax embedding. Embryo blocks and composite blocks of postnatal organs were sectioned at ∼ 8 μ m, and the sections taken up onto slides treated with 3-aminopropyltriethoxy-silane (BDH) (Rentrop et al. 1986). Following dewaxing and rehydration, the sections were digested for 10 min in 0· 25 mg ml-1 pronase E (Sigma, autodigested for 4h at 37 °C in H2O, freeze-dried in single-use samples) in 50 mm-Tris–HCl pH7·5/5 rrtm-EDTA at 37°C, and postfixed for 20 min in 4% paraformaldehyde at room temperature. In some experiments, this step was omitted and in others the digestion was performed with Proteinase K (Boehringer) at concentrations of 1, 4 or 8 μg ml-1 for 30min at 37°C. The slides were then dehydrated through graded alcohols and dried. As a negative control some sections were incubated for 2h at 37°C in 400 μg ml-1 RNase A (Sigma) and 3 μg ml-1 RNase T1 (Sigma) immediately after the protease digestion step.

In situ hybridization

Hybridization buffer (30 μl) containing 100 ng ml-1 probe was pipetted onto the sections, a 22 mm × 32 mm coverslip applied and the slides incubated at 50°C for 16h. The hybridization buffer comprised 50 % deionized formamide (Fisons), 0·3m-NaCl, 10mm-Tris–HCl pH7·5, 10mm-Na2HPO4 pH6·8, 5 mm-EDTA, 0·02% BSA, 0·02% Ficoll 400, 0 ·02% polyvinylpyrrolidone, 1 mg ml-1 yeast RNA, 10mm-DTT and 10% dextran sulphate (Pharmacia). Slides from each block were probed either with labelled RNA complementary to the mRNA (generated with T7 polymerase) or with labelled RNA homologous to the mRNA (negative control, generated with T3 polymerase).

The coverslips were removed and the slides washed for 4h in three changes of 700 ml of hybridization buffer without RNA, DTT or dextran sulphate at 50°C. They were then incubated at 37°C for 1h in 500 ml of 0·5m-NaCl/ 10 mm-Tris–HCl (pH 7·5)/1 mm-EDTA to which was added 100 μg ml-1 RNase A (Sigma, made up at 50 mg ml-1 in water immediately prior to use, boiled for 1 min then added to the buffer). The slides were then given two 30 min washes in 2×SSC at 65 °C, and a final 30 min wash in 0· l×SSC at 65 °C (all washes were performed in shaking water baths). The slides were then dehydrated through graded alcohols containing 0·3M-NH4AC, and dried.

Autoradiography

The slides were coated with a liquid photographic emulsion melted at 45°C (Ilford K5 diluted 10g/5·9ml H2O/0·1ml glycerol), allowed to drain and dry upright for 18h. This extended drying time prior to desiccation appears to reduce ‘edge labelling’, a common autoradiographic artefact. The slides were exposed desiccated at 4°C for 7–21 days. Slides were developed by sequential immersion in 160 g l-1 Kodak D19 (4min), 1 % acetic acid (1 min), 150 g 1”-1 Kodak Unifix (4 min) and were finally washed in water. All the solutions were precooled to a uniform temperature of 15 °C. The slides were allowed to come to room temperature and washed in water for at least 30min, then stained with haematoxylin and eosin, dehydrated, cleared and coverslips applied.

Antibody staining

The preparation and purification of the rabbit anti-laminin antiserum has been described previously (Forster et al. 1984). Slides were dewaxed, rehydrated and digested in 0·025 % pepsin in 0·01 M-HCI at 37°C for 45 min. They were then incubated with the antiserum and washed in PBS prior to detection by the anti-rabbit three-stage peroxidase/antiperoxidase complex method. Localization was demonstrated by the diaminobenzidine–hydrogen peroxidase reaction (Forster et al. 1984). Negative control slides were treated as above but the anti-laminin antiserum was replaced with preimmune serum from the animals used to raise the laminin antibody.

In situ hybridization (ISH)

Using the ISH method described, consistent results with good labelling and low background were obtained from different batches of probe. No spurious hybridization was observed with the negative control probe. RNase pretreatment abolished specific cellular labelling with the positive probe. In most cases, the most intense labelling was seen following pronase E digestion (Fig. 1A,B). A lower level of labelling was usually observed in the absence of Pronase treatment (not shown). Proteinase K digestion at 1 μg ml-1 for 30min at 37°C gave lower levels of labelling (Fig. 1C), and with increasing enzyme concentration the signal was progressively lost (Fig. 1D). This phenomenon was most striking in large embryo blocks and less apparent in blocks of organ fragments, possibly due to greater crosslinking between protein and mRNA in smaller tissue pieces during fixation.

Fig. 1.

Effect of protease digestion on ISH signal. (A) Bright-field photograph of detail from a whole 12-to 13-day embryo section shows heart (h), which contains fetal red cells (example arrowed) and liver (l). (B–D) Dark-field microscopy of three sections showing the same region of the block. (B) Section digested with 125 μg ml-1 Pronase E (37°C, 10min), intense labelling of myocardium (m) and liver capsule (c); unlabelled fetal red blood cells (example arrowed) are birefringent under the dark-field illumination. (C) Section digested with lμg ml-1 Proteinase K (37°C, 30 min), labelling in the myocardium (m) is reduced to below that from the unlabelled birefringent red cells (example arrowed), and labelling is abolished in the liver capsule (c). (D) Section digested with 4 μ g ml-1 Proteiriase K (37°C, 30min), labelling in myocardium (m) is virtually reduced to background level and only birefringent red cells (example arrowed) are visible. Bar, 0 · 1mm.

Fig. 1.

Effect of protease digestion on ISH signal. (A) Bright-field photograph of detail from a whole 12-to 13-day embryo section shows heart (h), which contains fetal red cells (example arrowed) and liver (l). (B–D) Dark-field microscopy of three sections showing the same region of the block. (B) Section digested with 125 μg ml-1 Pronase E (37°C, 10min), intense labelling of myocardium (m) and liver capsule (c); unlabelled fetal red blood cells (example arrowed) are birefringent under the dark-field illumination. (C) Section digested with lμg ml-1 Proteinase K (37°C, 30 min), labelling in the myocardium (m) is reduced to below that from the unlabelled birefringent red cells (example arrowed), and labelling is abolished in the liver capsule (c). (D) Section digested with 4 μ g ml-1 Proteiriase K (37°C, 30min), labelling in myocardium (m) is virtually reduced to background level and only birefringent red cells (example arrowed) are visible. Bar, 0 · 1mm.

Decidua, extraembryonic membranes and placenta

During the early postimplantation stage of mouse embryo development (up to ∼8 days post coitum) Reichert’s membrane showed intense staining for laminin protein and there was also a much fainter staining BM delineating the boundary between the embryonic ectoderm and mesoderm (Fig. 2A). ISH showed that the only cells of embryonic derivation that contained laminin mRNA at a detectable level were the parietal yolk-sac endoderm cells (Fig. 2B,C). The trophoblastic giant cells and decidual cells around the implantation site had no detectable laminin protein or laminin mRNA (Fig. 2A–C). In those areas of the mesometrial and antimesometrial endometrium in which decidualization of the stroma was actively occurring, laminin-containing blood vessel BMs were detected and, in those areas, the large decidual cells showed intense labelling for laminin mRNA (not shown). At the 10day stage, similar laminin staining of blood vessel BM was observed in both mesometrial and antimesometrial endometrium (Fig. 3A,D). In the decidual cells at the mesometrial aspect of the uterus, uniform, high-level, expression of laminin mRNA was detected (Fig. 3B,C) whereas mRNA expression in the antimesometrium (decidua capsularis) was restricted to a thin layer corresponding to the zone where recanalization of the uterus occurs (Fig. 3E,F; the basal zone, Stewart & Peel, 1978). The trophoblastic cells of the ectoplacental cone showed no protein or detectable mRNA, but the maternal cells bordering the cone stained cytoplasmically for laminin (Fig. 3G), and had the highest levels of laminin mRNA seen in maternal cells (Fig. 3H,I). Throughout the subsequent development of the extraembryonic membranes, the parietal endoderm and endodermal sinuses displayed laminin-rich BM (Reichert’s membrane; Fig. 4A) and expressed the highest levels of laminin mRNA, whereas the endoderm of the visceral yolk sac expressed little detectable laminin mRNA (Fig. 4B,C). Within the placental labyrinth the fetal capillaries had laminin-containing BMs (Fig. 4D). Laminin mRNA expression in this region was associated with small endothelial-like cells rather than with cells lining maternal blood spaces, which do not possess BMs (Fig. 4E,F). The spongiotrophoblast (junctional zone) showed no detectable laminin mRNA; however, within this area there were isolated islands of intensely labelled cells. The trophoblastic giant cells of the giant cell layer were surrounded by a laminin-containing BM and many, but not all, were intensely labelled following ISH, and showed a characteristic perinuclear distribution of silver grains. Accumulation of laminin protein at the junction of the maternal and fetal tissue was detected, and laminin mRNA was also present in maternal cells in this region (Fig. 4G–I). The results are summarized in Table 1.

Table 1.

Detection of laminin mRNA and protein in maternal and extraembryonic tissues

Detection of laminin mRNA and protein in maternal and extraembryonic tissues
Detection of laminin mRNA and protein in maternal and extraembryonic tissues
Fig. 2.

Implantation chamber 7 days post coitum. (A) Laminin immunoperoxidase staining, laminin protein is detected in Reichert’s membrane (filled arrows), between embryonic ectoderm and mesoderm (open arrows) and in blood vessel BM in maternal decidua (b). (B) A section from same block as A probed for laminin mRNA; (C) same field as B but with dark-field illumination to show silver grains in autoradiograph. Laminin mRNA is localized in parietal endoderm cells (examples arrowed), with no labelling in embryonic tissues or decidua. Bar, 50 μm.

Fig. 2.

Implantation chamber 7 days post coitum. (A) Laminin immunoperoxidase staining, laminin protein is detected in Reichert’s membrane (filled arrows), between embryonic ectoderm and mesoderm (open arrows) and in blood vessel BM in maternal decidua (b). (B) A section from same block as A probed for laminin mRNA; (C) same field as B but with dark-field illumination to show silver grains in autoradiograph. Laminin mRNA is localized in parietal endoderm cells (examples arrowed), with no labelling in embryonic tissues or decidua. Bar, 50 μm.

Fig. 3.

10-day uterus. (A) Mesometrial segment; laminin antibody staining is seen in uterine muscle fibres (m) and maternal blood vessel BM (example arrowed). There is no intracellular staining in decidual cells (d). (B) Bright-field photograph of in situ hybridization to the same region; silver grains can be seen over decidual cells (d) rather than blood vessel endothelium (example arrowed). (C) Same field as B viewed under dark-field illumination shows the extent of decidual labelling. (D) Antimesometrial aspect of uterus; laminin antibody staining shows similar pattern to A. (E,F) ISH to a section from the same region photographed under bright-field (E) and dark-field (F) illumination; detectable laminin mRNA expression is restricted to basal zone (b) also there is no labelling of muscle (m). (G) Ectoplacental cone region, laminin antibody stain; maternal cells (s) in the immediate vicinity of the trophoblast have detectable cytoplasmic laminin; maternal cells remote from the trophoblast (d) lack cytoplasmic immunoreactivity as do the trophoblast cells (t). (H,I) ISH to a section from the same region photographed under bright-(H) and dark-(I) field illumination, shows the expression of laminin mRNA is more intense in maternal cells (s) associated with the trophoblast (t) than in those remote from it (d). There is no detectable expression of laminin mRNA in trophoblastic cells. Bar, 0·1 mm.

Fig. 3.

10-day uterus. (A) Mesometrial segment; laminin antibody staining is seen in uterine muscle fibres (m) and maternal blood vessel BM (example arrowed). There is no intracellular staining in decidual cells (d). (B) Bright-field photograph of in situ hybridization to the same region; silver grains can be seen over decidual cells (d) rather than blood vessel endothelium (example arrowed). (C) Same field as B viewed under dark-field illumination shows the extent of decidual labelling. (D) Antimesometrial aspect of uterus; laminin antibody staining shows similar pattern to A. (E,F) ISH to a section from the same region photographed under bright-field (E) and dark-field (F) illumination; detectable laminin mRNA expression is restricted to basal zone (b) also there is no labelling of muscle (m). (G) Ectoplacental cone region, laminin antibody stain; maternal cells (s) in the immediate vicinity of the trophoblast have detectable cytoplasmic laminin; maternal cells remote from the trophoblast (d) lack cytoplasmic immunoreactivity as do the trophoblast cells (t). (H,I) ISH to a section from the same region photographed under bright-(H) and dark-(I) field illumination, shows the expression of laminin mRNA is more intense in maternal cells (s) associated with the trophoblast (t) than in those remote from it (d). There is no detectable expression of laminin mRNA in trophoblastic cells. Bar, 0·1 mm.

Fig. 4.

Placenta 19 days. (A) Laminin antibody staining of Reichert’s membrane (r) and endodermal sinus BM (e). (B,C) ISH to similar sections photographed using bright-(B) or dark-(C) field illumination show high levels of laminin mRNA expression in parietal endoderm cells (p) (these have become detached during processing) and in endodermal sinus cells (e). Expression is much lower in visceral yolk sac (v). (D) Laminin antibody staining of placental labyrinth demonstrates laminin in the BM of fetal vessels (example arrowed). (E,F) ISH to sections of labyrinth photographed under bright-(E) and dark-(F) field illumination reveals labelling associated with small endothelial-like cells (examples arrowed). (G) Laminin antibody staining of junctional zone region shows accumulation of laminin at the junction of the fetal and maternal tissue (ma). Laminin is also present in BM associated with giant cells (g) (arrowed). There is little laminin-containing BM in spongiotrophoblast (s). (H,I) ISH of similar region of junctional zone photographed under bright-(H) and dark-(I) field illumination; laminin mRNA expression is seen in maternal tissue especially at the fetal/matemal boundary (ma); expression is also detectable in some giant cells (g). Compared to labyrinth (1) there is little detectable laminin expression in spongiotrophoblast (s) except for a few groups of cells (example arrowed). Bar, 0·lmm.

Fig. 4.

Placenta 19 days. (A) Laminin antibody staining of Reichert’s membrane (r) and endodermal sinus BM (e). (B,C) ISH to similar sections photographed using bright-(B) or dark-(C) field illumination show high levels of laminin mRNA expression in parietal endoderm cells (p) (these have become detached during processing) and in endodermal sinus cells (e). Expression is much lower in visceral yolk sac (v). (D) Laminin antibody staining of placental labyrinth demonstrates laminin in the BM of fetal vessels (example arrowed). (E,F) ISH to sections of labyrinth photographed under bright-(E) and dark-(F) field illumination reveals labelling associated with small endothelial-like cells (examples arrowed). (G) Laminin antibody staining of junctional zone region shows accumulation of laminin at the junction of the fetal and maternal tissue (ma). Laminin is also present in BM associated with giant cells (g) (arrowed). There is little laminin-containing BM in spongiotrophoblast (s). (H,I) ISH of similar region of junctional zone photographed under bright-(H) and dark-(I) field illumination; laminin mRNA expression is seen in maternal tissue especially at the fetal/matemal boundary (ma); expression is also detectable in some giant cells (g). Compared to labyrinth (1) there is little detectable laminin expression in spongiotrophoblast (s) except for a few groups of cells (example arrowed). Bar, 0·lmm.

Embryonic and postnatal tissues and organs

Central nervous system (CNS), autonomic nervous system (ANS)

In the 10-day-old embryo, the external limiting membrane surrounding the neural tube stained strongly for laminin protein. However, laminin mRNA was not detectable in the neuroepithelium by ISH. At later stages, as the cerebral vesicles developed, the external limiting membrane of the cerebral vesicles and neural tube remained positive for laminin protein (Fig. 5A) and groups of cells in the adjacent pia mater showed detectable expression of laminin mRNA, particularly around blood vessels (Fig. 5B,C). Within the neural tissue itself, laminin protein was detected around small blood vessels and ISH demonstrated a few positive cells in these areas, but no generalized laminin mRNA expression was detectable in neuronal tissue (Fig. 5B,C). The internal limiting membrane was weakly positive for laminin protein, but mRNA was not reliably demonstrated in this region (not shown).

Fig. 5.

12-day embryo. (A) Dorsal root ganglion stained by the laminin antibody, the emerging fibres (f) are stained, as are some cells within the ganglion (g). The ganglion is surrounded by a laminin-containing membrane (m). (B,C) ISH of a similar section photographed under (B) bright-field and (C) dark-field illumination shows uniform labelling for laminin mRNA over emerging fibres (f). The cells within the pia mater (arrowed) bordering the neural tube (nt) are also labelled but the neural tissue is unlabelled. (D) High-power detail of the ganglion shows uniform distribution of silver grains over emerging fibres but heterogeneous labelling of cells within the ganglion (g). 19-day embryo nerve outgrowth. (E) Nerve fibres (n) are stained by the anti-laminin antibody. (F,G) ISH of a similar section photographed under bright-(F) and dark-(G) field illumination shows laminin mRNA expression in cells associated with the nerve fibres (n). Bars: A – C, 0 · l mm; D, 20 μ m; E – G, 0 · 2mm.

Fig. 5.

12-day embryo. (A) Dorsal root ganglion stained by the laminin antibody, the emerging fibres (f) are stained, as are some cells within the ganglion (g). The ganglion is surrounded by a laminin-containing membrane (m). (B,C) ISH of a similar section photographed under (B) bright-field and (C) dark-field illumination shows uniform labelling for laminin mRNA over emerging fibres (f). The cells within the pia mater (arrowed) bordering the neural tube (nt) are also labelled but the neural tissue is unlabelled. (D) High-power detail of the ganglion shows uniform distribution of silver grains over emerging fibres but heterogeneous labelling of cells within the ganglion (g). 19-day embryo nerve outgrowth. (E) Nerve fibres (n) are stained by the anti-laminin antibody. (F,G) ISH of a similar section photographed under bright-(F) and dark-(G) field illumination shows laminin mRNA expression in cells associated with the nerve fibres (n). Bars: A – C, 0 · l mm; D, 20 μ m; E – G, 0 · 2mm.

Dorsal root ganglia were surrounded by a laminin-rich capsule (Fig. 5A) and laminin protein, and mRNA was present in some cells within the ganglia (Fig. 5D). These cells were most probably the neuro-lemmal cells which are known to secrete laminin in vitro (Bunge & Bunge, 1985). The dorsal root fibres were also very strongly stained by the antibody (Fig. 5A) and laminin mRNA was present (Fig. 5B,C), probably in Schwann cells. An identical pattern of antibody staining and laminin mRNA expression was observed in ventral root ganglia and fibres (not shown). Throughout the embryo from the 12-day stage, nerve outgrowths could be seen to contain laminin and express laminin mRNA (Fig. 5E – G).

Gut

At 12 days the epithelium lining the small intestine had a laminin-containing BM and there was cytoplasmic reactivity and detectable mRNA in the mesenchymal cells immediately adjacent to the epithelium. The epithelium itself did not show cytoplasmic staining for laminin and ISH failed to demonstrate the presence of laminin mRNA (not shown). By 15 days until around the time of birth, the muscularis externa of the small intestine was clearly recognizable surrounding an inner mucosa which is covered by a simple columnar epithelium thrown up into villus ridges. The outer muscularis externa was strongly stained by the antibody and cells in the mucosa showed strong immunoreactivity, but the epithelium remained negative (Fig. 6A). ISH confirmed that laminin mRNA was present in the muscle and the lamina propria (Fig. 6B,C), particularly in cells subjacent to the epithelium (Fig. 6D). In tissues taken around the time of weaning (17 – 22 days after birth), the small bowel had assumed a near adult appearance with larger very cellular crypts and villi. Staining for laminin protein was restricted to the muscularis externa, muscularis mucosa and lamina propria cells subjacent to the epithelium (Fig. 6E). ISH revealed that laminin mRNA expression was strongest in cells in the muscularis externa, muscularis mucosa and in the lamina propria (Fig. 6F,G). Within the lamina propria the labelling was not entirely uniform and was highest in cells underlying the epithelium (Fig. 6H). In adult intestine, expression of laminin mRNA could not be reliably detected above background in any cell type in the small bowel by ISH. The same pattern of laminin protein staining and mRNA expression was seen during the development of the stomach and colon though labelling of the muscle and mesenchyme was much stronger, making it difficult to assess possible epithelial labelling because of ‘crossfire’ effects from adjacent heavily labelled mesenchyme cells. During the period from birth to weaning, cells in the mesentery and in connective tissue associated with pancreatic acini were also strongly labelled with ISH, though the acini themselves did not express laminin mRNA at a detectable level (not shown).

Fig. 6.

19-day embryonic small intestine. (A) The laminin antibody stains the muscularis externa (m) and cells in the lamina propria (l) underlying the epithelium (e). (B,C) ISH of a similar section photographed under bright-(B) and dark-(C) field illumination shows laminin mRNA expression throughout the muscularis externa (m) and the lamina propria (l) but not in the epithelium (e). (D) Higher magnification of a section through a small villus shows silver grains over cytoplasm of lamina propria cells. 20 day postnatal small intestine. (E) Immunocytochemistry reveals less prominent staining of muscularis externa (m) for laminin than in A, but intense staining in the lamina propria (l), particularly in the region adjacent to the epithelium (e). (F,G) ISH to a similar section photographed under bright-(F) and dark-(G) field illumination shows laminin mRNA expression in muscularis externa (m) and in the lamina propria (l) but not in the epithelium (e). Bright refractile spots (examples arrowed) corresponding to basal epithelial cells in the crypts are Paneth cell granules. (H) High-magnification photograph of the crypt/villus junction shows the heterogeneity of labelling within the lamina propria, the cells most closely associated with the epithelium, probably fibroblasts, are those most often seen to be labelled. Bars: A – C.E – G, 0 · 1 mm; D,H, 25 μ m.

Fig. 6.

19-day embryonic small intestine. (A) The laminin antibody stains the muscularis externa (m) and cells in the lamina propria (l) underlying the epithelium (e). (B,C) ISH of a similar section photographed under bright-(B) and dark-(C) field illumination shows laminin mRNA expression throughout the muscularis externa (m) and the lamina propria (l) but not in the epithelium (e). (D) Higher magnification of a section through a small villus shows silver grains over cytoplasm of lamina propria cells. 20 day postnatal small intestine. (E) Immunocytochemistry reveals less prominent staining of muscularis externa (m) for laminin than in A, but intense staining in the lamina propria (l), particularly in the region adjacent to the epithelium (e). (F,G) ISH to a similar section photographed under bright-(F) and dark-(G) field illumination shows laminin mRNA expression in muscularis externa (m) and in the lamina propria (l) but not in the epithelium (e). Bright refractile spots (examples arrowed) corresponding to basal epithelial cells in the crypts are Paneth cell granules. (H) High-magnification photograph of the crypt/villus junction shows the heterogeneity of labelling within the lamina propria, the cells most closely associated with the epithelium, probably fibroblasts, are those most often seen to be labelled. Bars: A – C.E – G, 0 · 1 mm; D,H, 25 μ m.

Kidney

In the developing kidney, both the nephrogenic vesicles, derived from the metanephric blastema and the ampullae, derived from the ureteric buds, were attached to a laminin-containing BM (Fig. 7A) and expressed high levels of laminin mRNA (Fig. 7B,C). Subdivisions of the ureteric bud which ultimately differentiate into the collecting ducts and the undifferentiated metanephrogenic cap had no detectable laminin mRNA (Fig. 7B,C). Despite the intensity of ISH labelling in the cells in differentiating metanephric nephron, there was little cytoplasmic staining for protein, and the only cytoplasmic immunoreactivity was seen in cells between the developing nephrons (Fig. 7A). This pattern of ISH labelling of laminin mRNA in the developing nephrons, Bowman’s capsules and blood vessels continued throughout embryonic life. Postnatally the kidney undergoes considerable further development. From birth to around 17 days, laminin mRNA expression was detected in the loops of Henle and around the collecting ducts of the medullary rays, but expression of laminin mRNA was undetectable by ISH in the proximal and distal convoluted tubules (Fig. 7D – F). The intensity of protein staining in BMs and the level of laminin mRNA expression in all labelled structures declined with time until after weaning when no laminin mRNA expression in any cell type could be convincingly demonstrated above background (Fig. 7G – I). During the early phase of this postnatal stage, some intracellular immunoreactivity was seen in cell types with demonstrable laminin mRNA, but this declined with time until by weaning laminin was only detectable in BM.

Fig. 7.

15-day embryo kidney. (A) Laminin immunocytochemistry, ampullae and nephrogenic vesicles have laminin BM but no intracellular staining (examples arrowed). (B,C) ISH to a similar section photographed under bright-(B) and dark-(C) field illumination reveals that laminin mRNA expression is seen in ampullae (a), and nephrogenic vesicles (v), but not in metanephric blastema (mb), or in other subdivisions of the ureteric bud (u). 8-day postnatal kidney. (D) The laminin antibody stains Bowman’s capsule and all tubular and blood vessel BM strongly. (E,F) ISH to a similar section photographed under bright-(E) and dark-(F) field illumination shows some laminin mRNA expression still present, principally in cells in the medullary rays (mr), rather than in glomeruli (g), or in proximal or distal convoluted tubules (ct). 20-day postnatal kidney. (G) Laminin antibody staining in all BM, except in glomerular capillaries (g), is much reduced. (H,I) ISH shows no detectable mRNA in any cell type. Bars: A,D,G, 50 μ m; B,C,E,F,H,I, 0 · 1mm.

Fig. 7.

15-day embryo kidney. (A) Laminin immunocytochemistry, ampullae and nephrogenic vesicles have laminin BM but no intracellular staining (examples arrowed). (B,C) ISH to a similar section photographed under bright-(B) and dark-(C) field illumination reveals that laminin mRNA expression is seen in ampullae (a), and nephrogenic vesicles (v), but not in metanephric blastema (mb), or in other subdivisions of the ureteric bud (u). 8-day postnatal kidney. (D) The laminin antibody stains Bowman’s capsule and all tubular and blood vessel BM strongly. (E,F) ISH to a similar section photographed under bright-(E) and dark-(F) field illumination shows some laminin mRNA expression still present, principally in cells in the medullary rays (mr), rather than in glomeruli (g), or in proximal or distal convoluted tubules (ct). 20-day postnatal kidney. (G) Laminin antibody staining in all BM, except in glomerular capillaries (g), is much reduced. (H,I) ISH shows no detectable mRNA in any cell type. Bars: A,D,G, 50 μ m; B,C,E,F,H,I, 0 · 1mm.

Liver

Throughout embryonic and postnatal development, the liver parenchyma showed little laminin immunoreactivity and no detectable laminin mRNA. At around 8 days after birth there appeared to be some laminin mRNA present and close examination revealed that the most strongly labelled cells were not hepatocytes but sinusoidal and Kupfer cells, which also had detectable laminin protein within their cytoplasm, and cells associated with developing bile ducts. In the capsular region and in the periportal connective tissue, there was some detectable protein and mRNA which was most apparent immediately prior to birth and during the first 15 days of postnatal development. Following this postnatal period, the expression of laminin mRNA and the intensity of protein staining declined in these regions and, in the adult liver, no immunoreactivity, except around blood vessels, and no ISH labelling above background in any cell type were detectable (not shown).

Heart

Throughout its embryonic development, the heart showed uniform labelling by ISH and a laminin-containing BM was laid down around the myocardial cells (Fig. 8A – C). In the immediate postnatal period, no expression of laminin mRNA was detected over the cardiac muscle. However, the connective tissue cells associated with the cardiac vasculature strongly expressed laminin mRNA (Fig. 8D – F).

Fig. 8.

12-to 13-day embryonic heart. (A) Intracellular laminin immunoreactivity is seen in the developing, myocardium. (B,C) ISH to a similar section photographed under bright-(B) and dark-(C) field illumination shows uniform expression of laminin mRNA in the developing heart. 15-day postnatal heart (D) laminin protein is restricted to the BM around cardiac muscle fibres and blood vessels. (E,F) ISH to a similar section photographed under bright-(E) and dark-(F) field illumination reveals that cardiac muscle fibres have no detectable laminin mRNA but connective tissue cells (examples arrowed) are heavily labelled. Bar, 0 · 1 mm.

Fig. 8.

12-to 13-day embryonic heart. (A) Intracellular laminin immunoreactivity is seen in the developing, myocardium. (B,C) ISH to a similar section photographed under bright-(B) and dark-(C) field illumination shows uniform expression of laminin mRNA in the developing heart. 15-day postnatal heart (D) laminin protein is restricted to the BM around cardiac muscle fibres and blood vessels. (E,F) ISH to a similar section photographed under bright-(E) and dark-(F) field illumination reveals that cardiac muscle fibres have no detectable laminin mRNA but connective tissue cells (examples arrowed) are heavily labelled. Bar, 0 · 1 mm.

In situ hybridization methodology

After comparing numerous protocols and suggested pretreatments, the simplified method described here was found to give the best results in terms of sensitivity, reproducibility, signal-to-noise ratio and histological preservation of the tissue. The method can be used on tissue fixed by other mild crosslinking compounds by varying the protease digestion step. Most importantly it can be employed on conventional formalin-fixed material routinely used in pathology departments.

Modulating the protease digestion step is the key to sensitivity in this particular system and is very dependent on the length of fixation time and type of fixative employed. The size of the blocks used and the differing penetration rates of fixative in different tissues also appears to influence results, hence our use of differing protease digestion schedules to ensure optimum mRNA availability in all tissues used. We found that Pronase E digestion at 125 μg ml-1 for 10 min at 37°C gave the best signal in large blocks (Fig. 1) compared to digestion with Proteinase K at 1 μ g ml-1 for 30 min at 37 °C, however in smaller blocks similar signals were obtained with both enzymes. All the results presented in this study are based on Pronase E-digested slides.

The technique reliably labelled the same cell types in duplicate experiments, however the sensitivity of the technique is not such that it will identify cells expressing very low levels of mRNA. For example, in adult rat kidney a low level of laminin mRNA expression can just be detected by Northern blotting (Dr N. Samani, personal communication) but ISH does not give convincing signals. Similarly, we have demonstrated very low levels of laminin mRNA in adult rat intestinal mucosa by dot-blot hybridization (data not shown) but could not reliably detect cellular localization by ISH to sections of rat intestine using the same 35S-labelled RNA probe. Thus we cannot exclude the possibility that cells negative by ISH express laminin mRNA at very low level.

The specificity of the probe for laminin was assessed in a number of ways, (a) A transcript of the second strand of the cDNA (identical to the mRNA sequence) labelled to the same specific activity and used at the same concentration on parallel tissue sections failed to hybridize, (b) The probe hybridized strongly and specifically to tumour cells in sections of the EHS sarcoma, from which the laminin protein was originally isolated (Timpl et al. 1979) and not to surrounding host stromal tissue, (c) The probe hybridized strongly and specifically to parietal endoderm cells from which the laminin cDNA was isolated (Barlow et al. 1984). (d) Other single-stranded RNA probes (exon 3 of c-myc, c-erb-B2, fibronectin) did not show the same pattern of hybridization when used on sections from blocks used in this study, (e) Prehybridization digestion with 300 μ g ml-1 RNase A for 2 h at 37 °C reduced labelling in all positive cells to background levels.

Laminin gene expression during embryogenesis

During the early postimplantation phase of embryonic development, detectable laminin production was restricted to the parietal yolk-sac endoderm and to maternal decidual cells which are responsible for laminin production in the mesometrial and antimesometrial endometrium. At the 10-day stage, the detectable expression of laminin in the antimesometrial endometrium was restricted to the zone where recanalization of the uterine lumen occurs, raising the possibility that laminin expression may be involved in guiding the growing uterine epithelium.

During the early development of the placenta, the invading trophoblast showed no demonstrable laminin expression and the intense mRNA expression seen at the junction of the ectoplacental cone with the endometrium was again maternally derived. Out-growth of cells from blastocysts cultured in vitro on laminin matrices has been reported (Armant et al. 1986a,b), and it is thus possible that the maternal cells secrete extracellular matrix components which guide trophoblast migration. The parietal endoderm cells retained their intense expression of laminin mRNA throughout embryonic life, reflecting the increasing thickness of Reichert’s membrane and also, possibly, turnover of BM. In the mature placenta, other cells that express laminin mRNA were found associated with fetal capillaries in the labyrinth and in small clusters in the junctional zone; in addition, trophoblastic giant cells at the interface with the maternal tissue also express laminin mRNA.

In the embryonic tissues, we detected immunoreactivity for laminin in peripheral nerves; it was present in cells associated with the ventral root fibres and in a small proportion of cells in the dorsal root ganglia. ISH revealed mRNA in a proportion of cells in dorsal root ganglia, in cells associated with ventral root fibres and in all peripheral nerve outgrowths. The cells producing this laminin are probably Schwann cells which are known to secrete several BM components in vitro including laminin (Bunge & Bunge, 1985). Schwann-cell-derived extracellular matrix has been shown to promote neurite outgrowth in vitro and guide axon regeneration in vivo (Ide et al. 1985). In common with Sternberg & Kimber (1986), we found the external limiting membrane of the neural tube contained laminin, but in the 10-day embryo we could not identify the cells responsible for mRNA synthesis by ISH. At later stages of embryonic development, the laminin associated with the outer limiting membrane of the cerebral vesicles and spinal cord appeared to be synthesized by the adjacent pia and not by the neuroepithelium.

The cells responsible for the secretion of epithelial basement membrane laminin in the intestine have not, to our knowledge, been identified reliably. In this study, we showed by ISH that laminin was produced by mesenchymally derived lamina propria cells and as the gut matures in postnatal life labelling is restricted to cells subjacent to the epithelium, i.e. those corresponding to the fibroblast sheath in the adult gut. The mRNA extracted from mucosal scraping of adult rat small intestine (which contains mRNA from epithelium and lamina propria, including fibroblast sheath cells) has low, but detectable, levels of laminin mRNA when assessed by dot-blot hybridization using a 35S-labelled single-stranded RNA probe for laminin. However, the same probe failed to show specific cellular localization in sections of rat intestine by ISH (data not shown). This suggests that despite the high level of cell migration and turnover in the small intestine (reviewed by Wright, 1980) production of new BM is low. Our failure to reliably demonstrate mRNA in fibroblasts of the adult gut in both rats and mice by ISH is probably due to inadequate sensitivity of the ISH technique. Situations where BM production is likely to be enhanced in the adult, e.g. during the adaptive response to intestinal resection (Senior et al. 1986) or during chronic mechanical trauma (Senior et al. 1985), could be used to study this further.

The role of laminin in the development of the kidney has been investigated in vitro and in vivo by Ekblom (Ekblom et al. 1980; Sariola & Ekblom, 1985). These studies revealed that laminin protein was detectable in kidney mesenchyme culture only in cells destined to become epithelial. Our ISH results show that there is no detectable expression of laminin mRNA in the metanephric blastemal cap but differentiation of these mesenchymal cells into epithelial nephrogenic vesicles was accompanied by expression of laminin mRNA at a similar level to that found in the epithelial ureteric buds. In this zone where fusion of the ureteric buds with the metanephrogenic cells occurs, laminin mRNA expression was seen in the ampullae cells, but at this stage cells from the region of the ureteric bud from which the collecting ducts develop have no detectable laminin mRNA. After birth, laminin mRNA became detectable in elongating loops of Henle and collecting ducts, and was reduced to below detectable levels around weaning. In contrast the glomeruli, proximal and distal convoluted tubules had little detectable laminin mRNA during this postnatal period. We did not observe detectable cytoplasmic immunoreactivity for laminin by peroxidase staining in ampullae or nephrogenic vesicles during the embryonic stage of renal development. However, we are confident that the ISH labelling of these structures is not artifactual because; (a) the negative control probe failed to hybridize to these regions, (b) prehybridization RNase digestion abolished the signal. It is possible that the protein is incorporated into BM without detectable intracellular accumulation.

The liver did not show any appreciable immuno-reactivity or labelling except in the capsular region until after birth when there was labelling in the perilobular connective tissue, particularly that associated with the developing biliary tree. There was also a burst of laminin transcription in the Kupfer and sinusoidal cells which was most intense at 8 days postnatally, but as expression and immunoreactivity are lost just prior to weaning, its significance is unclear.

The heart showed uniform labelling by ISH for laminin over the developing myocardium until after birth, suggesting that the myocardial cells are responsible for producing their BM laminin. The embryonic mouse heart contains extractable laminin protein (Coughlin et al. 1986) and we demonstrated intracellular immunoreactivity. The expression of laminin mRNA in cardiac muscle cells declined to below detectable levels around birth. In postnatal life, laminin expression was restricted to connective tissue cells and a similar situation seems to occur in smooth muscle such as that of the colon and stomach.

The most intense levels of mRNA expression were seen in tissues undergoing growth, reorganization or migration, that is those laying down new basement membrane or extracellular matrix. For example, in relevant somites of the 10-day embryo, migrating fusiform sclerotome cells stain with the laminin antibody whereas the as yet undifferentiated epithelial dermamyotome cells are negative. At a later stage, migratory myotome cells become positive (data not shown).

In all mature tissues examined, laminin mRNA expression was below the level detectable by ISH, e.g. the smooth muscle of the uterine wall has a laminin-rich BM surrounding the muscle fibres, but no labelling by ISH was ever detected. Similarly in the adult small intestine even with its high rate of epithelial cell turnover and migration (reviewed by Wright, 1980), laminin mRNA could not be reliably demonstrated by ISH, suggesting that the turnover of this BM component is low. In some instances, the correlation between apparent cytoplasmic protein staining and levels of labelling seen with ISH was poor, e.g. maternal decidual cells in both the meso-metrial and antimesometrial segments of the uterus and developing nephrons in the embryonic kidney all showed intense labelling by ISH, but little demonstrable intracellular protein. These results illustrate the value of ISH as a method for identifying cells producing a protein which is not accumulated cyto-plasmically.

In the developing kidney, the expression of detectable levels of laminin mRNA is restricted to the epithelial cells, and it appears that these cells are responsible for the synthesis of basement membrane laminin. In contrast no convincing labelling of epithelial cells in the intestine could be demonstrated at any time during prenatal or postnatal life, indicating that in this tissue the BM laminin appears to be the product of the adjacent mesenchyme.

This study has demonstrated the usefulness of combining in situ hybridization and immunocytochemistry in studying the production and distribution of proteins during development. In the case of laminin, it is clear that either mesenchyme or epithelia produce this BM component depending on the tissue, and that expression of mRNA occurs concomitantly with growth, elongation and migration. In view of the wide distribution of cells expressing laminin mRNA in the developing embryo, in extraembryonic and maternal decidual tissues, further studies on the distribution of both the 67 and 140 × 103Mr receptor are likely to be of value in elucidating the role of the extracellular matrix in development.

This study was supported by a grant from the Cancer Research Campaign. We are indebted to Dr B. M. Hogan (NIMR, Mill Hill) for providing the laminin cDNA, to Dr P. Holland (NIMR, Mill Hill) and Dr H. Pringle (Department of Pathology, Leicester) for advice on ISH technique, Dr S. Peel and Dr I. Stewart (Department of Human Morphology, Southampton) for assistance in the interpretation of the matemal/placental tissue results, and to Professor D. Moffat (Department of Anatomy, Cardiff) for assistance in the interpretation of some of the embryonic kidney results and to the technical staff of the University Department of Pathology for tissue sectioning and immunocytochemistry.

Amenta
,
P. S.
,
Gay
,
S.
,
Vaheri
,
A.
&
Martinez-Hernandez
,
A.
(
1986
).
The extracellular matrix is an integrated unit: ultrastructural localization of collagen types, I, III, IV, V, VI, fibronectin and laminin in human term placenta
.
Collagen Rel. Res.
6
,
125
152
.
Armant
,
D. R.
,
Kaplan
,
H. A.
&
Lennarz
,
W. J.
(
1986a
).
Fibronectin and laminin promote in vitro attachment and outgrowth of mouse blastocysts
.
Devi Biol.
116
,
519
523
.
Armant
,
D. R.
,
Kaplan
,
H. A.
,
Mover
,
H.
&
Lennarz
,
W. J.
(
1986b
).
The effect of hexapeptides on attachment and outgrowth of mouse blastocysts cultured in vitro: evidence for the involvement of the cell recognition tripeptide Arg-Gly-Asp
.
Proc. natn. Acad. Sci. U.S.A.
83
,
6751
6755
.
Barlow
,
D. P.
,
Green
,
N. M.
,
Kurkinen
,
M.
&
Hogan
,
B. L. M.
(
1984
).
Sequencing of laminin B chain cDNAs reveals C-terminal regions of coiled-coil alphahelix
.
EMBO J.
3
,
2355
2362
.
Bozyczko
,
D.
&
Horwitz
,
A. F.
(
1986
).
The participation of a putative cell surface receptor for laminin and fibronectin in peripheral neurite extension
.
J. Neurosci.
6
,
1241
1251
.
Buck
,
C. A.
&
Horwitz
,
A. F.
(
1987
).
Cell surface receptors for extracellular matrix molecules
. In
Annual Review of Cell Biology, vol. 3
(ed.
G. E.
Palade
,
B. M.
Alberts
&
J. M.
Spudich
), pp.
179
205
.
Pal Alto, California
:
Annual Reviews Inc
.
Bunge
,
M. B.
&
Bunge
,
R. P.
(
1985
).
Source and function of basal lamina in peripheral nerve
. In.
Basement Membranes,
(ed. S. Shibata). pp. 363-371. Amsterdam: Elsevier Science
Publishers
B.V.
Cohen
,
J.
,
Burne
,
J. F.
,
Mckinlay
,
C.
&
Winter
,
J.
(
1987
).
The role of laminin and the laminin/fibronectin receptor complex in the outgrowth of retinal ganglion cell axons
.
Devi Biol.
122
,
407
418
.
Coughlin
,
M. D.
,
Grover
,
A. K.
&
Jung
,
C. Y.
(
1986
).
Determination of the molecular weight of neuronectin, a conditioned medium-derived, substrate-binding neurite-extension factor: comparison with laminin using radiation-inactivation analysis
.
J. Neurosci.
6
,
1553
1559
.
Cox
,
K. H.
,
Deleon
,
D. V.
,
Angerer
,
L. M.
&
Angerer
,
R. C.
(
1984
).
Detection of mRNAs in sea urchin embryos by in situ hybridization using asymmetric RNA probes
.
Devi Biol.
101
,
485
502
.
Davis
,
G. E.
,
VarÓn
,
S.
,
Engvall
,
E.
&
Manthorpe
,
M.
(
1985
).
Substratum-binding neurite-promoting factors: relationships to laminin
.
Trends Neurosci.
8
,
528
532
.
Dohrmann
,
U.
,
Edgar
,
D.
,
Sendtner
,
M.
&
Thoenen
,
H.
(
1986
).
Muscle-derived factors that support survival and promote fibre outgrowth from embryonic chick spinal motor neurons in culture
.
Devi Biol.
118
,
209
221
.
Dziadek
,
M.
&
Timpl
,
R.
(
1985
).
Expression of nidogen and laminin in basement membranes during mouse embryogenesis and in teratocarcinoma cells
.
Devi Biol. III, 372-382
.
Ekblom
,
P.
,
Alitalo
,
K.
,
Vaheri
,
A.
,
Timpl
,
R.
&
Saxen
,
L.
(
1980
).
Induction of a basement membrane glycoprotein in embryonic kidney: possible role of laminin in morphogenesis
.
Proc. natn. Acad. Sci. U.S.A.
77
,
485
489
.
Engel
,
J.
,
Odermatt
,
E.
&
Engel
,
A.
(
1981
).
Shapes, domain organizations and flexibility of laminin and fibronectin, two multifunctional proteins of the extracellular matrix
.
J. molec. Biol.
150
,
97
120
.
Forster
,
S. J.
,
Talbot
,
I. C.
&
Critchley
,
D. R.
(
1984
).
Laminin and fibronectin in rectal adenocarcinoma: relationship to tumour grade, stage and metastasis
.
Br. J. Cancer.
50
,
51
61
.
Grabel
,
L.
&
Watts
,
T. D.
(
1987
).
The role of extracellular matrix in the migration and differentiation of parietal endoderm from teratocarcinoma embryoid bodies
.
J. Cell Biol.
105
,
441
448
.
Graf
,
J.
,
Iwamoto
,
Y.
,
Sasaki
,
M.
,
Martin
,
G. R.
,
Kleinman
,
H. K.
,
Robey
,
F. A.
&
Yamada
,
Y.
(
1987
).
Identification of an amino acid sequence in laminin mediating cell attachment, chemotaxis and receptor binding
.
Cell
48
,
989
996
.
Horwitz
,
A.
,
Duggan
,
K.
,
Greggs
,
R.
,
Decker
,
C.
&
Buck
,
C.
(
1985
).
The cell substrate attachment (CSAT) antigen has properties of a receptor for laminin and fibronectin
.
J. Cell Biol.
101
,
2134
2144
.
Ide
,
C.
,
Tohyama
,
K.
,
Onodera
,
S.
,
Nitatori
,
T.
&
Osawa
,
T.
(
1985
).
Schwann cell basal laminae and central nervous regeneration: observations on the early stages of regeneration
. In
Basement Membranes
(ed.
S.
Shibata
), pp.
353
361
.
Amsterdam
:
Elsevier Science Publishers B.V
.
Kurosawa
,
K.
,
Isemura
,
M.
,
Yamaguchi
,
Y.
,
Yosizawa
,
Z.
,
Furuyama
,
T.
,
Yoshinaga
,
K.
&
Ishii
,
T.
(
1985
).
Changes in distribution of connective tissue components of human placentae with maturation
.
Tohoku J. exp. Med.
147
,
261
265
.
Longo
,
F. M.
,
Hayman
,
E. G.
,
Davis
,
G. E.
,
Ruoslahti
,
E.
,
Engvall
,
E.
,
Manthorpe
,
M.
&
VarÓn
,
S.
(
1984
).
Neurite-promoting factors and extracellular matrix components accumulating in vivo within nerve regeneration chambers
.
Brain Res.
309
,
105
117
.
Maniatis
,
T.
,
Fritsch
,
E.
&
Sambrook
,
J.
(
1982
).
Molecular Cloning: A Laboratory Handbook.
New York
:
Cold Spring Harbour Laboratory
.
Martin
,
G. R.
&
Timpl
,
R.
(
1987
).
Laminin and other basement membrane components
. In
Annual Review of Cell Biology,
vol.
3
(ed.
G. E.
Palade
,
B. M.
Alberts
&
J. M.
Spudich
), pp.
57
85
.
Pal Alto, California
:
Annual Reviews Inc
.
Mcclay
,
D. R.
&
Ettensohn
,
C. A.
(
1987
).
Cell adhesion and morphogenesis
. In
Annual Review of Cell Biology, vol. 3
(ed.
G. E.
Palade
,
B. M.
Alberts
&
J.
Ohno
,
M.
,
Martinez-Hernandez
,
A.
,
Ohno
,
N.
&
Kefalides
,
N. A.
(
1985
).
Comparative study of laminin found in normal placental membrane with laminin of neoplastic origin
. In
Basement Membranes
(ed.
S.
Shibata
), pp.
3
-
11
. Amsterdam: Elsevier Science Publishers.
Ott
,
U.
,
Odermatt
,
E.
,
Engel
,
J.
,
Furthmayer
,
H.
&
Timpl
,
R.
(
1982
).
Protease resistance and conformation of laminin
.
Eur. J. Biochem.
123
,
63
72
.
Pytela
,
R.
,
Pierschbacher
,
M. D.
&
Ruoslahti
,
E.
(
1985
).
Identification and isolation of a 140kd cell surface glycoprotein with properties expected of a fibronectin receptor
.
Cell
40
,
191
198
.
Rao
,
N. C.
,
Barsky
,
S. H.
,
Terranova
,
V. P.
&
Liotta
,
L. A.
(
1983
).
Isolation of a tumor cell laminin receptor
.
Biochem. biophys. Res. Commun.
Ill
,
804
808
.
Rao
,
N. C.
,
Margulies
,
I. M. K.
,
Tralka
,
T. S.
,
Terranova
,
V. P.
,
Madri
,
J. A.
&
Liotta
,
L. A.
(
1982
).
Isolation of a subunit of laminin and its role in molecular structure and tumor cell attachment
.
J. biol. Chem.
257
,
9740
9744
.
Reh
,
T. A.
,
Nagy
,
T.
&
Gretton
,
H.
(
1987
).
Retinal pigmented epithelial cells induced to transdifferentiate to neurons by laminin
.
Nature, Lond.
330
,
68
71
.
Rentrop
,
M.
,
Knap
,
B.
,
Winter
,
H.
&
Schweizer
,
J.
(
1986
).
Aminoalkylsilane-treated glass slides as support for in situ hybridization of keratin cDNAs to frozen tissue sections under varying fixation and pretreatment conditions
.
Histochem. J.
18
,
271
276
.
Rogers
,
S. L.
,
Letourneau
,
P. C.
,
Palm
,
S. L.
,
Mccarthy
,
J.
&
Furcht
,
L. T.
(
1983
).
Neurite extension by peripheral and central nervous system neurons in response to substratum-bound fibronectin and laminin
.
Devi Biol.
98
,
212
220
.
Sariola
,
H.
&
Ekblom
,
P.
(
1985
).
Extracellular matrix in Wilms tumor and the embryonic kidney
. In
Basement Membranes
(ed.
S.
Shibata
), pp.
407
—414. Amsterdam: Elsevier Science Publishers B.V.
Senior
,
P. V.
,
Pritchett
,
C. J.
,
Sunter
,
J. P.
,
Appleton
,
D. R.
&
Watson
,
A. J.
(
1985
).
Transplantation of a segment of ileum to the external abdominal wall: an animal model of intestinal mucosal hyperplasia
.
J. Path.
146
,
39
49
.
Senior
,
P. V.
,
Pritchett
,
C. J.
,
Sunter
,
J. P.
,
Appleton
,
D. R.
&
Watson
,
A. J.
(
1986
).
A cytokinetic study of the small-intestinal and colonic mucosa after 70 % resection of the small intestine
.
Cell Tiss. Kinet.
19
,
491
502
.
Sternberg
,
J.
&
Kimber
,
S. J.
(
1986
).
Distribution of fibronectin, laminin and entactin in the environment of migrating neural crest cells in early mouse embryos
.
J. Embryol. exp. Morph.
91
,
267
282
.
Stewart
,
I.
&
Peel
,
S.
(
1978
).
The differentiation of the decidua and the distribution of metrial gland cells in the pregnant mouse uterus
.
Cell Tissue Res.
187
,
167
-
179
. -
Timpl
,
R.
,
Rohde
,
H.
,
Robey
,
P. G.
,
Rennard
,
S. L
,
Foidart
,
J.M.
&
Martin
,
G. R.
(
1979
).
Laminin - a glycoprotein from basement membranes
.
J. biol. Chem.
254
,
9933
9937
.
Tuckett
,
F.
&
Morriss-Kay
,
G. M.
(
1986
).
The distribution of fibronectin laminin and entactin in the neurulating rat embryo studied by indirect immunofluorescence
.
J. Embryol. exp. Morph.
94
,
95
112
.
Wright
,
N. A.
(
1980
).
Cell proliferation in normal experimental animals
. In
Cell Proliferation in the Gastrointestinal Tract
(ed.
D. R.
Appleton
,
J. P.
Sunter
&
A. J.
Watson
), pp.
3
-
26
. Tunbridge Wells: Pitman Medical.