Expression of the two mannose 6-phosphate receptors is spatially and temporally different during mouse embryogenesis

Mannose 6-phosphate receptors (MPRs) mediate the targeting of newly synthesized soluble lysosomal enzymes. The receptors recognize mannose 6-phosphate residues which are specifically added to lysosomal enzymes during their passage through the Golgi apparatus. Receptor ligand complexes migrate from the trans Golgi network to prelysosomal compartments, where the complexes dissociate due to acidic pH. The ligands are delivered to lysosomes, while the MPR can recycle to the Golgi apparatus (Kornfeld and Mellman,1989) .

Two MPRs are known, both of which mediate the transport of newly synthesized mannose 6-phosphate-containing ligands from the Golgi apparatus to prelysosomal structures of the endocytic pathway. Furthermore, the larger of the two receptors with an apparent M_T of 300 ×10³ (MPR300) mediates the endocjlosis of mannose 6-phosphate-containing ligands, while the smaller of the two receptors with an apparent M_r of 46 ×10³ (MPR46) mediates the secretion of part of its ligands. The two receptors bind essentially the same mannose 6-phosphate-containing ligands although with different affinities and pH optima (Dahms et al., 1989; Chao et al., 1990). The insulin-like growth factor (IGF) II is another ligand of mammalian MPR300 (Morgan et al., 1987). The physiological function of IGFII binding to MPR300 is unknown. Most of the metabolic effects of IGFII can be explained by its binding to the IGFI receptor (Czech, 1989; Roth, 1988). While IGFII mediated coupling of the MPR300 to a G₁-protein suggests a function in transmembrane signalling (Murayama et al., 1990) other studies indicate a function of MPR300 in controlling the level of IGFII by its uptake and degradation (Czech, 1989; Haig and Graham, 1991) .

With the exception of a few tumor-derived cell lines that lack the MPR300 (Gabel et al., 1983; Mainferme et al., 1985) all mammalian cell lines and tissues examined contain both receptors. However, the concentrations and the relative ratios of the two receptors vary among several human cell lines and tissues by about one order of magnitude (Wenk et al., 1991).

Why do cells express two different MPRs? The most likely reason is that the two receptors have different or at least partly different functions, e.g. binding of IGFII. MPR300 has acquired the IGFII binding site fairly recently in evolution, since in non-mammalian species, such as chicken and frog, MPR300 lacks the IGFII binding site (Canfield and Kornfeld, 1989; Clairmont and Czech, 1989). It is not clear whether the acquisition of IGFII binding or the need for other functional differences has provided the evolutionary force for the development and expression of two MPRs.

In an attempt to further compare the properties of the two MPRs we have investigated their expression during mouse embryonic development from d9.5 to dl7.5. To our surprise we observed an almost complementary pattern of expression of the two receptors.

Mice, strain NMRI, were mated overnight. Females showing vaginal plugs on the next morning were kept separately and the beginning of gestation was defined as midnight. The embryos were sectioned, washed in ice cold IxPBS (0.13 M NaCl, 7 mM Na₂HPO₄,3 mM NaH₂PO₄ 2H₂O, pH 7.0), and blotted on a paper towel. For preparation of RNA, embryos were subsequently homogenized. For cryosectioning they were immediately frozen at the surface of dry ice, embedded in tissue-tec freeze medium (Reichert-Jung) and stored at −70°C for up to several weeks. Cryosectioning was done with the cryostat Modell 2700 Frigocut (Reichert-Jung).

Total RNA was prepared according to the protocol of Chirgwin et al. (1979). Poly(A)⁺ RNA was isolated as described by Gonda et al. (1982).

RNA probes for Northern blot analysis and in situ hybridization were obtained by in vitro transcription of receptor gene fragments cloned into the pGeml vector (Promega). The MPR46 mouse cDNA EcoRI fragment was 1397 bp in size and contained 99 bp of 5 ′ untranslated region, the entire coding region and 360 bp of 3 ′ non-translated sequences (Koster et al., 1991). The MPR300 mouse cDNA EcoRI fragment was 1263 bp in size (EMBL accession number X60389) and encoded a part of the luminal domain of the receptor corresponding to nt 5041 to 6303 of the human MPR300 gene (Oshima et al., 1988). Antisense and sense RNA probes were transcribed from a linearized vector using T7 RNA polymerase. For Northern blot analysis (Maniatis et al., 1982) RNA was labelled with [³²P]UTP, for in situ hybridization with [³⁵S]UTP using the in vitro transcription protocol of Holland (1986). The transcripts were precipitated twice with ammonium acetate to remove free nucleotides. Before hybridization of sections, riboprobes were hydrolyzed to an average size of 150 bases (Cox et al., 1984).

In situ hybridization was performed according to the protocol of Holland (1986) with the following modifications. Embryos were not fixed before sectioning. Sections were mounted on coated slides. For coating, the slides were cleaned in 10% Extran (Merck, Darmstadt, FRG) overnight, rinsed with warm water for 2 hours, washed three times with diethylpyrocarbonate (DEPC)-treated water and dried for 2 hours at 120°C. The slides were coated (2% aminopropyltriethoxysilan in acetone, 2 × acetone, 2 × DEPC-treated water, 30 seconds each), dried at 45°C, and stored at room temperature for up to several weeks.

The pretreatment with pronase was omitted. The ³⁵S-labelled riboprobes were diluted in hybridization buffer [300 mM NaCl, 10 mMTris-HCl pH 7.5,10 mM sodium phosphate buffer pH 6.8, 5 mM EDTA pH 8, l × Denhardt’s solution, 10% dextran sulfate, 100 mM DTT, 50% deionized formamide, 3 μg/ μl tRNA (E. coli), 10 μM α-thio-UTP, 0.5 μg/ μl heparine] to a specific activity of 5 ×10⁴ counts per minute/ μl. Hybridization was performed at 45°C.

The sections were washed 4 ×30 minutes in 50% formamide, 2 ×SSC, 10 mM )5-mercaptoethanol at 45°C, digested with 8 μg pancreatic RNAse A/ml RNAse buffer (0.5 M NaCl, 10 mM Tris-HCl pH 7.5,1 mM EDTA) for 30 minutes at 37°C and washed again 4 ×30 minutes in 2 ×SSC, 10 mM β-mercaptoethanol at 60°C, 2 ×30 minutes in 0.1 ×SSC, 10 mM β-mercaptoethanol at 60°C and 2 ×5 minutes in 0.l ×SSC at room temperature.

Following dehydration in graded ethanol containing 0.3 M ammonium acetate, sections were covered with NTB 2 photoemulsion (Kodak) and exposed for 2 to 3 weeks. Slides were developed, stained with cresyl violet and embedded with Eukitt (Kindler, Freiburg, FRG).

The probes for the MPR46 and MPR300 hybridize to specific RNA species of mouse Ltk ⁻ cells as shown by the Northern blot analysis (Fig. 1). The MPR46 riboprobe detects a main 2.3 kb transcript and two minor RNA species of 5 kb (not visible in Fig. 1) and 1.3 kb. The MPR300 riboprobe recognizes a unique RNA of about 10 kb. MPR46 and MPR300 transcripts of the same size are detectable in RNA from mouse embryos from d11 to dl7 and in adult mouse organs (shown for brain in Fig. 2). While the relative amount of MPR46 RNA is nearly constant from d11 to dl7 that of MPR300 RNA increases about 2.5-fold from d11 to dl3 and then remains fairly constant up to d17.

Mid- and parasagittal sections of embryos from d9.5 to d17.5 were analyzed. High levels of MPR46 transcription are detectable for the first time at dll.5 in liver, while MPR300 transcripts are detectable already at d9.5 in the heart and in umbilical vessels (Fig. 3). This pattern changes to a more complex distribution towards the later stages of embryogenesis (Fig. 3 and Fig. 4).

At higher magnification than shown in Fig. 3 it becomes apparent that the signal for MPR46 in liver from dl3.5 onwards originates mainly from cell aggregates surrounded by erythrocytes which give no signal. These aggregates are assumed to represent blood forming foci. In addition heavily labelled single cells morphologically identifiable as megakaryocytes are seen in liver from d15.5 onwards. The intervening patches of parenchmyal liver cells give much weaker signals.

Next to liver, MPR46 RNA is most abundant in the thymus (from d13.5 onwards) and in the upper intestinal tract (from d15.5 onwards). In the thymus the silver grains are initially distributed homogenously. At d17.5 they are concentrated in the subcapsular region which is enriched in large lymphoid cells and in the central part of the thymus above round cell patches.

Weaker MPR46 RNA signals are seen in the lung (d13.5), the somites and sklerotomal bodies (d11.5 to d13.5), the pancreas (d13.5), the submaxillary and sublingual glands (d15.5) and the skin (d15.5). The expression in the central nervous system is weak and has a maximum between d11.5 to d13.5.

At d17.5 the MPR46 expression reaches a complex pattern (Fig. 4A). Table 1 summarizes the tissues, in which MPR46 transcripts are detectable. Among the tissues that start to show prominent MPR46 labelling at dl7.5 are brown adipose tissue, thyroid, tooth anlage, hypophysis, adrenal gland, urogenital and lower intestinal tract, bone marrow and zones of ossification (see below).

At all stages of embryonic development the most prominent signals for MPR300 were seen in the cardiovascular system. Lower levels of MPR300 transcripts are detectable in the somites (d9.5), intersomitic clefts (dll.5) and in the liver close to the capsular surface and around blood vessels (d11.5). A strong labelling of the mandible becomes apparent at d11.5. At later stages Meckel’s cartilage (d13.5) and tongue (dl5.5) show prominent signals for MPR300. The tissues that express MPR300 transcripts at dl7.5 (Fig. 4B) are listed in Table 1. Strong labelling is now apparent in the intestinal tract, in smooth and striated muscles, the perichondrium/periosteum and zones of ossification in the vertebrae and the skull.

During organogenesis an almost mutually exclusive pattern of MPR46 and MPR300 expression was observed in liver, thymus, lung, brown adipose tissue, thyroid, submaxillary and sublingual glands, pituitary gland (all positive for MPR46), tongue and cardiovascular system (positive for MPR300). Even in tissues that express both receptors the expression of MPR46 and MPR300 can often be ascribed to different cell types. For example in the urinary bladder (Fig. 4) expression of MPR46 is restricted to the epithelial layer, while MPR300 transcripts are most abundant in the muscular wall. Further examples are shown in Fig. 5A-D. In the skin (Fig. 5A) the expression of MPR46 is concentrated in the ectodermally derived stratum germinativum and hair follicles, while MPR300 expression is most prominent in mesodermally derived layers of condensed connective tissue below the epidermis.

In the intestinal tract (shown for the small intestine in Fig. 5B) the expression of MPR46 and MPR300 is highest in the mucosal layer. MPR46 is most frequent in the epithelia of the forming crypts with significantly lower labelling of the epithelia covering the villi. MPR300 expression is scarce in the crypts, uniformly high in the epithelial layer of the villi and intermediate in the smooth muscle layer.

At zones of enchondral ossification MPR300 transcripts are abundant in the ossifying tissue, while MPR46 transcripts are absent or - if present - restricted to single cells. In the surrounding perichondrium/periosteum a strong ubiquitous expression of MPR300 is found, while the expression of MPR46 is restricted to single cells of unknown nature, which show a massive labelling (see Fig. 5C). In the clavicle (Fig. 4), which is one of the sites of early bone marrow formation, the bone marrow is heavily labelled for MPR46 while MPR300 transcripts are not seen. In the cerebellum at dl7.5 (Fig. 5D) the external and internal granular layer are positive for MPR46 while strong MPR300 RNA signals appear above the pia mater and the Purkinje cells.

The unexpected finding of this analysis of MPR46 and MPR300 expression during mouse development by in situ hybridization is the spatially and temporally differential expression of the two receptors. Even at late stages of embryogenesis, where MPR transcripts are detectable in most tissues, the expression pattern of the two MPRs is almost complementary. In several tissues that express both MPRs higher resolution showed that expression occurred either in different cell types or at different developmental stages of a cell type (e.g. intestinal mucosa, Fig. 5B). This differential expression pattern strongly argues for specific, yet unidentified functions of the two MPRs during organogenesis in mouse.

Tissues that are active in hemopoiesis and T-cell differentiation were the first that expressed high levels of MPR46 RNA. The MPR46 transcripts became detectable in liver (dll.5), thymus (dl3.5) and bone marrow (dl7.5) at the developmental stage when these tissues begin their function in hemopoiesis and blood cell differentiation. Hemopoiesis begins in liver at dll and in bone marrow at dl6, while the thymus is formed at dl2 and involved in T-cell maturation and selection from dl5 onwards (Rugh, 1991; Cristanti et al., 1986). Clearly it will be of interest to analyse the expression, distribution and function of MPR46 in hemopoietic and accessory cells of the bone marrow and thymus to obtain insight in the specific functions of MPR46 in these cells.

At later stages of embryonic development (d15.5-d17.5) MPR46 transcripts are abundant in tissues of endodermal origin (e.g. epithelia of the respiratory, digestive and urogenital tracts, salivary glands, thyroidea, pancreas), but are also found in a few derivatives of the mesoderm (e.g. kidney) and ectoderm (e.g. epidermis). No information is available about the protein level of MPR46 in any of the embryonic tissues.

The expression of the MPR300 during the pre- and the postimplantation period of mouse (Harvey and Kaye, 1991) and rat (Senior et al., 1990; Tollefsen et al., 1989) has been the subject of several studies. Transcripts of the MPR300 are produced only from the maternal chromosome, but not from the paternal (Barlow et al., 1991) and transcription is observed as early as the two cell stage of mouse development (Harvey and Kaye, 1991). Using in situ hybridization of developing rat, MPR300 transcripts were first detected in the extraembryonic membranes. In the embryo proper high levels of expression were observed during organogenesis and histodifferentation in mesoderm-derived tissues such as the cardiovascular system, developing muscle, perichondrium and periosteum and a few endoderm-derived tissues, such as the respiratory endoderm (Senior et al., 1990). This pattern agrees with that in mouse observed in this study. While in rat none of the ectoderm-derived tissues showed MPR300 expression, in mouse MPR300 transcripts were clearly detectable in the cerebellum, plexus choroideus and ganglion coeliacum. Hence MPR300 is expressed in derivatives of all three germ layers as is MPR46.

For a number of tissues high expression of MPR300 transcripts correlates with a high expression of IGFII transcripts (Senior et al., 1990; Tollefsen et al., 1989; Ohlsson et al., 1989). The coordinately regulated expression has suggested a role of the receptor in a paracrine/autocrine function of IGFII. There are however exceptions, and the most notable is liver. The liver parenchyma is one of the major sites of IGFII expression during embryonic development (Beck et al., 1987; Stylianopoulou et al., 1988), but expresses only low levels of MPR300 as shown in this and other studies (Senior et al., 1990).

For the cardiovascular system it is known that the high levels of MPR300 transcripts are paralleled by high levels of immunoreactive receptor (Senior et al., 1990; Tollefsen et al., 1989; Sklar et al., 1989). The functional importance of MPR300 for the cardiovascular system is suggested by the following observation. Mice that maternally inherit a deletion of the Tme locus and hence lack MPR300 expression become highly oedematous and die at d15 of embryogenesis (Barlow et al., 1991). The development of severe oedems points to a malfunction of the cardiovascular system, which from d11 onwards maintains a circulation of blood and hence is important for the provision of embryonic tissues with nutrients and the disposal of metabolic waste.

The most pertinent question raised by the present study certainly concerns the role of MPR46 and MPR300 during organogenesis and histodifferentiation. The few data available for the frequency of the two receptors in adult tissues (Wenk et al., 1991) suggest that differential expression patterns are also maintained in the postnatal period and this assumption is supported by preliminary data of in situ hybridization of adult mouse tissues (U. Matzner, unpublished). The additional function of the MPR300 as a receptor for IGFII provides a plausible explanation for the existence of two MPRs only if the evolution of MPR46 is correlated with the acquisition of the IGFII binding site by MPR300. Otherwise yet unknown receptor properties or other known functions that are specific for either type of MPR such as endocytosis and secretion of mannose 6-phosphate-containing ligands, which include growth factors (Kovacina et al., 1989; Dennis et al., 1991; Lee and Nathans 1988), may have necessitated the evolution of two types of MPRs and their differential expression during development.

This work was supported by the Deutsche Forschungs-gemeinschaft (SFB 236) and the Fonds der Chemischen Industrie.