Advertisement
Research Article
Peripheral lymphangiogenesis in mice depends on ectodermal connexin-26 (Gjb2)
Nikolai Dicke, Nicole Pielensticker, Joachim Degen, Julia Hecker, Oliver Tress, Tobias Bald, Alexandra Gellhaus, Elke Winterhager, Klaus Willecke
Journal of Cell Science 2011 124: 2806-2815; doi: 10.1242/jcs.084186

Summary

In order to study the specific function of connexin-26 (Cx26, also known as gap junction beta-2 protein; Gjb2), we generated knockin mice that expressed either a floxed lacZ reporter or, after Cre-mediated deletion, connexin-32 (Cx32)-coding DNA, both driven by the endogenous Cx26 promoter. Heterozygous Cx26knock-inCx32 (Cx26KICx32) embryos developed normally until embryonic day 14.5 but died before birth with severe lymphedemas. Although the jugular lymph sacs were normally developed, these embryos had a strongly reduced dermal lymphatic capillary network. By analyses of β-galactosidase reporter protein expression and lymphatic or blood endothelial-specific marker proteins, we demonstrated that Cx26 expression is temporally closely linked to lymphangiogenesis. No obvious phenotypic abnormalities were observed in Cx26KICx32 mice when Cre-mediated recombination was directed to mesenchyme or blood endothelium using the Prx1-Cre or Tie2-Cre mouse strains, respectively. By contrast, keratin-5-Cre-mediated replacement of Cx26 with Cx32 or deletion of both Cx26 alleles revealed severe lymphedemas similar to the general Cx26KICx32 phenotype. Thus, conditional ablation of Cx26 (loss of function) in ectoderm leads to partial disruption of lymphatic capillaries and embryonic death. We conclude that appropriate development of dermal lymphatic vessels in mice is dependent on the expression of Cx26 in the ectoderm.

Introduction

Cellular interactions are necessary for the organization of cells into multilayered tissues or organs during development and adulthood. Connexin-26 (Cx26, also known as gap junction beta-2 protein; Gjb2) and connexin-32 (Cx32/Gjb1) are members of the connexin family that codes for 20 different proteins in the mouse genome (Söhl and Willecke, 2003). Connexins oligomerize to hexameric connexons (hemichannels). Docking and opening of two connexons result in conduits that directly connect the cytoplasm of adjacent cells for diffusional exchange of ions and metabolites. These gap junction channels facilitate the diffusional exchange of small molecules up to 1.8 kDa molecular mass (Neijssen et al., 2005). Ablation of a connexin isoform (knockout or null mutation), certain point mutated connexin isoforms or exchange of different connexin isoforms (knockin; KI) can cause strong developmental defects and embryonic death (Krüger et al., 2000; Dobrowolski et al., 2009; Frank et al., 2010). In hepatocytes, Cx26 and Cx32 proteins are largely colocalized and occur together in gap junctions (Nicholson et al., 1987). Homozygous Cx26 knockout (KO) mice are early embryonically lethal [embryonic day (E) 9.5–10.5], because Cx26 is an essential subunit protein of gap junction channels for the exchange of metabolites between maternal and embryonic blood circuits in the placental labyrinth (Gabriel et al., 1998). We wanted to investigate whether Cx32 can functionally replace Cx26 during development and in adult liver. Surprisingly, we found that the monoallelic exchange of Cx26 by Cx32 in mice did not impair placental function but led to embryonic death between E16.5 and 18.5. The mutant embryos exhibited severe lymphedemas, suggesting malformation of lymphatic vessels. Thus we decided to study this unexpected phenotype in more detail.

The lymphatic system mediates tissue fluid homeostasis, immune surveillance and fat absorption, and has been shown to be involved in the pathogenesis of several human diseases. Nearly every tissue or organ system is pervaded by lymphatic vessels, with the exception of the central nervous system, bone marrow, cartilage, cornea and epidermis (Mäkinen et al., 2007; Karpanen and Alitalo, 2008). In contrast to the circulatory cardiovascular system, the lymphatic system is a unidirectional structure, built up by blind-ending lymphatic capillaries, precollector vessels and larger connecting vessels, which are interrupted by lymph nodes. These larger lymphatics are finally connected via the thoracic duct and the right lymphatic duct to the venous blood vessel system (Tammela et al., 2005). The lymphatic system develops after the establishment of the cardiovascular system. In mammals it has been found that the whole lymphatic endothelium exclusively originates from a subpopulation of venous cells, which give rise to the total lymphatic endothelium by centrifugal sprouting (Srinivasan et al., 2007). In birds and tadpoles, the lymphatic vasculature is of dual origin: lymphatic endothelial cells arise by transdifferentiation from venous endothelial cells and by differentiation of mesenchymal lymphangioblasts (Wilting et al., 2006; Ny et al., 2005). Recent data suggest a dual origin of lymphatic endothelial cells also in mice (Buttler et al., 2006; Buttler et al., 2008).

Several proteins have been shown to fulfill specific functions during lymphangiogenesis. For detailed reviews see Tammela et al. (Tammela et al., 2005), Mäkinen et al. (Mäkinen et al., 2007), Tammela and Alitalo (Tammela and Alitalo, 2010) and Oliver and Srinivasan (Oliver and Srinivasan, 2010). The transcription factor Prox1 is expressed in a subpopulation of endothelial cells and has been shown to be essential for the first cells that are committed to the lymphatic endothelial cell lineage on E10. Mice in which Prox1 is ablated, fail to develop the first lymphatic structures, the jugular lymph sacs, and exhibit severe lymphedemas on E14.5 (Wigle and Oliver, 1999). Endothelial cells that are committed to the lymphatic lineage also express the lymphatic hyaluronan (HA) receptor (LYVE-1) and the vascular endothelial growth factor receptor 3 (VEGFR-3) (Banerji et al., 1999; Wigle et al., 2002). Furthermore, vascular endothelial growth factor C (VEGF-C), a ligand of VEGFR-3, is required for polarized sprouting of the first lymphatic endothelial cells from embryonic veins. The authors suggested that the paracrine VEGF-C signal constitutes a concentration gradient towards which the first lymphatic endothelial cells migrate to form jugular lymph sacs (Karkkainen et al., 2004). VEGF-C knockout mice also lack the jugular lymph sacs (Karkkainen et al., 2004). Inhibition of VEGF-C and VEGF-D by expression of soluble VEGFR-3 under control of the keratin 14 promoter led to moderate lymphedemas due to regression of the dermal lymphatic network (Mäkinen et al., 2001).

Less is known about connexin functions in lymphangiogenesis and lymph vessel integrity. Recently it has been shown that certain mutations in the human gap junction protein Cx47 can cause inherited human lymphedema (Ferrell et al., 2010; Ostergaard et al., 2011). In this study we investigated the essential role of ectodermally expressed Cx26 for peripheral lymphangiogenesis in mice.

Results

Generation of transgenic mice

We generated mice in which the coding region of Cx26 was replaced by a β-galactosidase (β-gal) coding region, flanked by loxP sites, preceded by a nuclear localization signal and followed by the coding region of Cx32 (Fig. 1A). The targeting vector was designed to express, under control of endogenous Cx26 regulatory elements, either lacZ- or, after Cre recombinase-mediated deletion, Cx32 coding DNA. Thus, we could analyze Cx26 expression by use of the lacZ reporter gene (in Cx26+/floxlacZ[Cx32] mice) or, alternatively, we were able to study the replacement of Cx26 with Cx32 (in Cx26+/Cx32 mice; Fig. 1A). Mice were genotyped by polymerase chain reaction (PCR; Fig. 1B). Southern blot hybridizations, using external and internal probes, demonstrated homologous recombination in mice (Fig. 1C). Transgenic Cx26+/floxlacZ[Cx32] mice were viable and did not show any phenotypic abnormalities. They were used to study reporter gene expression and for breeding with different Cre-recombinase-expressing mouse lines. As overview, all the mouse mutants analyzed in this study are summarized in supplementary material Table S1.

Fig. 1.
Fig. 1.

Generation of conditional Cx26knockinCx32 mice. (A) Scheme of homologous recombination of the targeting vector into the coding region of Cx26. The resulting transgenic allele (Cx26floxlacZ[Cx32]neo) includes a nuclear location signal (nls) fused to the lacZ coding DNA, which is flanked by loxP sites, followed by Cx32 coding DNA and a neomycin selection cassette, which is flanked by frt sites. Flp recombinase activity causes deletion of the neomycin selection cassette (Cx26floxlacZ[Cx32]). After Cre-mediated deletion of the lacZ reporter DNA, Cx32 is driven by the endogenous Cx26 promoter (Cx26Cx32neo or Cx26Cx32). (B) PCR products specific for wild-type and transgenic loci using DNA from embryonic yolk sac tissues. (C) Homologous recombination was confirmed by Southern blot hybridization of BglII-digested genomic DNA prepared from embryos (E15.5) using an external 5′ Cx26 probe and an internal Cx32 probe. Fragment sizes of each genotype are indicated in the scheme.

Heterozygous Cx26KICx32 embryos exhibit lymphedemas

For ubiquitous expression of Cx32 instead of Cx26, we crossed Cx26+/floxlacZ[Cx32] mice with phosphoglycerate kinase (PGK)-Cre-expressing mice (Lallemand et al., 1998), in order to obtain heterozygous Cx26KICx32 mice (Cx26+/Cx32). Surprisingly, we did not find any postnatal live mice of the Cx26+/Cx32 genotype. Hence, we focused on embryogenesis and found that Cx26+/Cx32 embryos died during development with severe malformations (Fig. 2). However, embryonic death was not due to malnutrition of the embryo caused by transplacental transport defects as previously found with Cx26KO embryos on E10.5 (Gabriel et al., 1998), because Cx26+/Cx32 embryos were not growth retarded and died later, between E16.5 and 18.5, as a result of fatal malformations. Until E14.5 no macroscopic malformations of Cx26+/Cx32 embryos were found (Fig. 2A,B) but mutants beyond E14.5 exhibited severe tissue swelling and agnathia (Fig. 2A–F). The inner organs developed normally until E16.5, but the surrounding mesenchymal tissues had undergone severe swelling (Fig. 2B,D,F). Higher magnification revealed big intercellular spaces between the mesenchymal cells relative to controls, pointing to accumulation of fluid in the matrix (Fig. 2F; supplementary material Fig. S1). The epidermis appeared disturbed in body regions next to strong accumulation of fluid, whereas in regions where less tissue swelling occurred it was stage specifically differentiated, e.g. in caudal regions (supplementary material Fig. S1). Because tissue fluid homeostasis is regulated and maintained by the vascular systems, we next concentrated on the cardiovascular system. Serial sections of Cx26+/Cx32 embryos showed normal development of the heart and surrounding blood vessels on E12.5 (supplementary material Fig. S2). To explain the subcutaneous swelling in the mesenchyme, we analyzed dermal blood vessel pattern and lymphatic vessel morphology in detail. Blood vessel morphology appeared macroscopically normal on E15.5 when accumulation of fluid had already been initiated (Fig. 3A–C). Moreover, a normal pattern of dermal blood capillaries was further demonstrated by immunohistochemical staining of E14.5 embryos, using the blood vessel endothelial-specific marker MECA-32 (Fig. 3D–G). Because we could not find any indication of cardiovascular defects, we tested whether components of the lymphatic vascular system were malformed. Therefore we analyzed the whole developing lymphatic vascular system of Cx26+/Cx32 embryos using lymphatic endothelial-specific marker proteins. Expression of LYVE-1, the earliest known marker of lymphatic endothelium (Tammela and Alitalo, 2010), was investigated in whole-mount embryos and embryo sections. In comparison with Cx26+/+ or Cx26+/floxlacZ[Cx32] embryos the peripheral lymphatic capillary network of Cx26+/Cx32 embryos was strongly reduced or even absent on E14.5 (Fig. 4A–D). At the same stage dermal blood vasculature appeared normal. The connecting network of dermal lymphatic capillaries was missing, and from E14.5 onwards the embryos developed lymphedemas. By contrast, Cx26+/Cx32 embryos showed normally developed early lymphatic anlagen, i.e. jugular lymph sacs, when compared with control Cx26+/floxlacZ[Cx32] embryos (Fig. 4E,F). Immunofluorescence analyses of the lymphatic endothelial-specific marker protein Prox1 on E15.5 also demonstrated the lack of dermal lymphatic vessels in Cx26+/Cx32 embryos, while in the deep embryo, Prox1-positive lymphatics were present (Fig. 4G–J). These findings suggest that the permutation affects development of peripheral lymphatics but not deeper parts of the lymphatic vessel system. Taken together, lethality of Cx26+/Cx32 embryos was not caused by a defective placental function, skin development or blood vessel morphology but by severe lymphedemas, which were the consequence of an impaired dermal lymphatic capillary network.

lacZ reporter gene expression during embryogenesis

Cx26 promoter activity, identified by β-gal staining, was not detected before E10.5. Between E10.5 and 17.5, Cx26+/floxlacZ[Cx32] embryos showed increasing lacZ expression in the developing skin of the embryo, in subcutaneous jugular regions, in the developing subarachnoidal space (Fig. 5A,B) and in the placenta (supplementary material Fig. S3).

The earliest β-gal activity was found on E10.5 in dorsolateral parts of the embryonic mesenchyme and in the regions of the branchial arches (Fig. 5A). From E10.5 to 13.5 it was present in the ectoderm in a highly regional manner, emerging more laterally than ventrally and dorsally in the embryonic body. Later, between E14.5 and 17.5, β-gal activity appeared more uniformly distributed in the epidermis (Fig. 5A). In addition to being present in embryonic skin, reporter gene expression was strongly abundant subcutaneously in the mesenchyme, from the jugular to caudal regions (E12.5–14.5) and in the developing subarachnoidal space (>E12.5; Fig. 5A,B). In the subarachnoidal space β-gal-positive cells were very closely associated to LYVE-1-positive cells (supplementary material Fig. S4).

Fig. 2.
Fig. 2.

Histological analyses of embryonic lethal Cx26+/Cx32 embryos. (A,B) Comparison of Cx26+/Cx32 embryos and control littermates (Cx26+/+). Cx26+/Cx32 embryos die between E16.5 and 18.5 showing accumulations of tissue fluid (lymphedemas; asterisk) and agnathia. Malformation of Cx26+/Cx32 embryos only become macroscopically visible after E14.5. (C–F) Paraffin sagittal sections of (C,E) control embryo (Cx26+/+) and (D,F) Cx26+/Cx32 littermate on E16.5, showing severe swelling of the subcutaneous mesenchyme (asterisk). Most organs seem to be normally developed, but the embryo exhibits severe lymphedemas. Scale bars: 1 mm (A–D); 250 μm (E,F).

Fig. 3.
Fig. 3.

Normal pattern of blood vasculature in Cx26+/Cx32 embryos. (A,B) Cx26+/Cx32 embryo on E15.5 compared with the control littermate (Cx26+/floxlacZ[Cx32]). The Cx26+/Cx32 embryo exhibits strong lymphedema and agnathia but the dermal blood vascular system is still intact. Blood circulated through the vessels, and the heart was still beating. (C) Magnification of the back of the Cx26+/Cx32 embryo (the arrow indicates a blood vessel). (D–G) Whole-mount immunohistochemical staining (E14.5) for the blood endothelial marker MECA-32. No differences in the dermal blood vessel network are evident in a Cx26+/Cx32 embryo and a control littermate (Cx26+/+). Scale bars: 1 mm (A,B); 500 μm (C); 250 μm (D,E); 125 μm (F,G).

Expression level of Cx26 and transgenic Cx32

Because it was known for other organs that the ablation of one connexin isoform can influence the level of another connexin isoform (Simon and McWhorter, 2003), we investigated whether the monoallelic expression of ectopic Cx32 influenced expression of Cx26.

Therefore we determined the amount of Cx26 in embryonic skin (Fig. 6A,B) and in placenta (Fig. 6C–E). In Cx26+/floxlacZ[Cx32] embryonic skin, the expression of the unaltered Cx26 allele was sufficient for the production of the normal level of Cx26, which was similar to that in Cx26+/+ controls. But Cx26 protein was significantly reduced in Cx26+/Cx32 embryonic skin compared with Cx26+/floxlacZ[Cx32] skin, although both genotypes were heterozygous for Cx26 (Fig. 6A,B). Because of the very low amount of Cx26 protein relative to the total protein content in embryonic skin lysates, we focused on Cx26 protein ratios and transcripts in the placenta. We chose the placenta, which in the mouse is of exclusive embryonic origin, because reporter gene analysis showed very strong expression in the labyrinth (cf. supplementary material Fig. S3). In comparison with Cx26+/+ placenta, Cx26+/floxlacZ[Cx32] and Cx26+/floxlacZ[Cx32]neo placentas showed nearly a 50% reduction in Cx26 protein (Fig. 6C,D). This is in contrast to Cx26+/floxlacZ[Cx32] embryonic skin, where monoallelic expression was sufficient for normal Cx26 expression. However, the reduced Cx26 in these placentas seemed to be sufficient for normal development of the mice, because they did not show any abnormalities. Remarkably, in comparison with the genotypes Cx26+/floxlacZ[Cx32] and Cx26+/floxlacZ[Cx32]neo, placentas of the genotypes Cx26+/Cx32 and Cx26+/Cx32neo exhibited a further significant decrease of Cx26 protein level (Fig. 6C,D). Thus in Cx26+/Cx32 embryonic skin as well as placentas, Cx26 was abnormally reduced compared with Cx26+/floxlacZ[Cx32] tissues. This finding was confirmed by a strong loss of Cx26 immunostaining and strong Cx32 immunosignals in the Cx26+/Cx32 placental labyrinth (supplementary material Fig. S5). On the transcript level, the genotypes Cx26+/floxlacZ[Cx32], Cx26+/floxlacZ[Cx32]neo, Cx26+/Cx32 and Cx26+/Cx32neo expressed comparable amounts of Cx26 mRNA but this was approximately half of the level found in Cx26+/+ placenta (Fig. 6E). Thus transcript levels were proportional to protein levels in Cx26+/floxlacZ[Cx32] and Cx26+/floxlacZ[Cx32]neo placentas, whereas in Cx26+/Cx32 and Cx26+/Cx32neo placentas the Cx26 protein levels were proportionally lower than the transcript levels. This finding indicates a posttranscriptional reduction of Cx26 in Cx26+/Cx32 placentas. As expected, transgenic Cx32 mRNA was only found in the genotypes Cx26+/Cx32 and Cx26+/Cx32neo, whereas no endogenous Cx32 mRNA was expressed in placentas (Fig. 6E). In summary, ectopic expression of Cx32 led to an abnormal reduction of the Cx26 protein level in Cx26+/Cx32 embryonic skin and placenta.

Cell-type-specific Cx26KICx32

The pattern of the reporter gene expression raised the question of which Cx26-expressing cell populations are involved in malformations and embryonic lethality of heterozygous Cx26KICx32 embryos. To discriminate between diverse cell populations we used three different transgenic mouse lines that expressed Cre recombinase under different promoter elements for the generation of heterozygous, cell-type-specific Cx26KICx32 mice.

Since Cx26 expression was found in different regions of the mesenchyme, we tested whether development of lymphatic capillaries is disturbed when Cre recombinase is active in mesenchymal cells in the genetic background of Cx26KICx32 mice. Therefore we mated female Cx26+/floxlacZ[Cx32] mice with male Prx1–Cre mice, in which Cre activity is directed to the mesenchyme of the limb buds, the interlimb flanks and craniofacial mesenchyme (Logan et al., 2002). The resulting Cx26+/floxlacZ[Cx32]:Prx1–Cre mice were viable and showed no obvious phenotypic abnormalities (data not shown). Because it has been reported that in mammals lymphatic endothelial cells originate from a subpopulation of venous endothelial cells, we investigated the effect in Cx26KICx32 mice when Cre expression is directed to blood endothelial cells. Hence we bred male Tie2–Cre mice (Constien et al., 2001) with female Cx26+/floxlacZ[Cx32] mice. The gene Tie2 codes for a receptor tyrosine kinase and its promoter is active in blood endothelial cells (Sato et al., 1993; Takakura et al., 1998). Interestingly, also the obtained Cx26+/floxlacZ[Cx32]:Tie2–Cre mice were viable and showed no obvious phenotypic abnormalities (data not shown). The fact that Cx26+/floxlacZ[Cx32]:Prx1–Cre and Cx26+/floxlacZ[Cx32]:Tie2–Cre mice did not exhibit lymphatic abnormalities suggests that neither the Prx1–Cre-positive mesenchyme nor the Tie2–Cre-positive endothelium are responsible for disturbed lymphangiogenesis of Cx26KICx32 embryos.

Fig. 4.
Fig. 4.

Malformation of dermal lymphatic capillaries in Cx26+/Cx32 embryos. (A–D) Whole-mount immunohistochemical staining (E14.5) for the lymphatic endothelial marker LYVE-1. Lymphatic capillaries in the Cx26+/Cx32 embryo are severely reduced, compared with the control littermate (Cx26+/+). (E,F) Immunohistochemical staining for LYVE-1 on transverse sections of embryos (E13.5). In comparison with the control embryo (Cx26+/floxlacZ[Cx32]), the Cx26+/Cx32 embryo has normally developed jugular lymph sacs (arrows) as indicated by the pattern of LYVE-1-positive cells. (G-J) Prox1-positive lymphatic endothelial cells in transverse sections of E15.5 embryos. (G,H) The Cx26+/Cx32 embryo has no Prox1-positive lymphatic vessels, whereas the control littermate (Cx26+/+) has several Prox1-positive lymphatic vessels within the dermis. (I,J) By contrast, in the interior of the embryo Prox1-positive lymphatics were detected in the Cx26+/Cx32 and the control littermate. nt, neural tube. Scale bars: 500 μm (A,B); 250 μm (C,D); 500 μm (E,F); 100 μm (G,H); 50 μm (I,J).

Next, we investigated whether the monoallelic ectodermal-directed replacement of Cx26 with Cx32 contributes to the lymphatic phenotype of Cx26KICx32 embryos. For this reason we bred male keratin 5–Cre (K5–Cre) mice (Ramirez et al., 2004) with female Cx26+/floxlacZ[Cx32] mice, in order to generate Cx26+/floxlacZ[Cx32]:K5–Cre embryos. Interestingly, Cx26+/floxlacZ[Cx32]:K5–Cre embryos exhibited a similar phenotype to Cx26+/Cx32 embryos, which was characterized by strong lymphedemas, agnathy and embryonic death (Fig. 7A). In order to determine the cell population in which K5–Cre-mediated exchange of Cx26 by Cx32 had occurred, we compared Cx26+/floxlacZ[Cx32]:K5–Cre embryos with Cx26+/+ and Cx26+/floxlacZ[Cx32] littermates. In Cx26+/floxlacZ[Cx32]:K5–Cre embryos between E12.5 and 15.5, the floxed lacZ reporter gene was deleted (Fig. 7B–D; supplementary material Fig. S6) and transgenic Cx32 was expressed only in ectodermal cells in which the K5 and Cx26 promoters were active (supplementary material Fig. S7). Cell lineage analysis of K5–Cre mice using the R26R reporter mouse strain (Soriano, 1999) also showed that K5–Cre was exclusively expressed in stratified epithelia as mentioned by Ramirez et al. (Ramirez et al., 2004). Cx32 was not expressed in either the β-gal-positive subcutaneous mesenchyme or the subarachnoidal space, indicating that these compartments are not involved in the disturbed lymphangiogenesis of the mutants (data not shown). However, transgenic Cx32 was abundantly expressed in the developing skin of Cx26+/floxlacZ[Cx32]:K5–Cre embryos, but not in the skin of control Cx26+/floxlacZ[Cx32] embryos (supplementary material Fig. S7). Our results demonstrated that between E12.5 and 15.5 the Cx26 and K5 promoters were active at the same time, in the same ectodermal cells. Although the monoallelic replacement of Cx26 with Cx32 did not take place in the entire Cx26+/floxlacZ[Cx32]:K5-Cre embryonic epidermis, the embryos developed severe lymphedemas. Because Cx26+/floxlacZ[Cx32]:K5-Cre embryos also exhibited the lymphatic defects observed in Cx26KICx32 embryos, we investigated whether dermal lymphatics were also absent. Whole-mount immunohistochemical staining for LYVE-1 demonstrated a strong reduction of the connecting lymphatic vasculature network in Cx26+/floxlacZ[Cx32]:K5-Cre embryos (Fig. 7E–H), similar to the ubiquitous Cx26KICx32 embryos.

These findings confirm that Cx26 in ectodermal cells is involved in peripheral lymphangiogenesis. Heterozygous replacement of Cx26 with Cx32, when restricted to ectodermal cells (mediated by K5–Cre), leads to malformations and strong decrease of the peripheral lymphatic network in Cx26+/floxlacZ[Cx32]:K5–Cre embryos, producing severe lymphedemas.

Expression analyses of growth factors and receptors relevant for lymphangiogenesis

To examine whether relevant factors for lymphangiogenesis were altered in Cx26+/floxlacZ[Cx32]:K5-Cre embryonic skin, we tested the expression of several factors on E14.5 by quantitative real time-polymerase chain reaction (qRT-PCR) analyses. We investigated the transcript levels of VEGF-C, VEGF-D, the corresponding receptors VEGFR-2, VEGFR-3 and the soluble form of VEGFR-2 (sVEGFR-2), which is known to antagonize VEGF-C action (Albuquerque et al., 2009). Furthermore, we tested neuropilin-2, a transmembrane receptor, whose interaction with VEGFR-3 mediates lymphatic vessel sprouting in response to VEGF-C (Xu et al., 2010). Surprisingly, in Cx26+/floxlacZ[Cx32]:K5-Cre embryonic skin, we found a significant upregulation of VEGF-C, while its receptors VEGFR-3, VEGFR-2 were significantly downregulated (Fig. 8). Furthermore, sVEGFR-2 was downregulated and no differences were detected in transcription level of neuropilin-2.

Fig. 5.
Fig. 5.

Cx26 expression in ectoderm, mesenchyme and subarachnoidal space. (A) Whole-mount Cx26+/floxlacZ[Cx32] embryos aged between E10.5 and 14.5 showing staining of β-gal. The earliest lacZ expression is detected on E10.5 in the branchial arches and in the dorsolateral body (arrowhead). From E12.5 onwards, embryos show strong staining subcutaneously (arrowhead) and in the developing skin. In E13.5 embryos, β-gal is present along the neural tube, i.e. in the developing subarachnoidal space (arrows) and subcutaneously (arrowhead). On E14.5, β-gal is more uniformly expressed in the skin. (B) Paraffin sagittal and transverse sections of Cx26+/floxlacZ[Cx32] embryos (E12.5) show that in addition to lacZ expression in the developing skin, it is also expressed in the subarachnoidal space (arrow) and in subcutaneous regions (arrowheads). Scale bars: 500 μm (A,B).

Ectodermal Cx26 function is essential for peripheral lymphangiogenesis

To determine whether the loss of Cx26 function, the gain of ectopic Cx32 function or both are responsible for the lymphatic phenotype of Cx26+/floxlacZ[Cx32]:K5-Cre embryos, we examined mice in which Cx26 was deleted in ectoderm (loss of function). Therefore, we bred male Cx26+/fl:K5-Cre mice with female Cx26fl/fl mice (Cohen-Salmon et al., 2002), in which the coding region of Cx26 was flanked by loxP sites (floxed). We obtained no living offspring of the genotype Cx26fl/fl:K5-Cre, because embryos developed lymphedemas and died between E16 and 18.5 (Fig. 9A,B). In comparison with Cx26+/Cx32 and Cx26+/flox[lacZ]Cx32:K5-Cre embryos, which developed severe lymphedemas, Cx26fl/fl:K5-Cre embryos exhibited a milder lymphatic phenotype. The extent of lymph accumulation in Cx26fl/fl:K5-Cre embryos was in some cases severe, but appeared milder in others. In accordance with this, the lymphatic capillary meshwork of Cx26fl/fl:K5-Cre embryos had been established but was much less interconnected compared with controls (Fig. 9C–F). In detail, the dermal lymphatic system of Cx26fl/fl:K5–Cre embryos appeared coarsely netted with enlarged diameter vessels, but the fine vessels were missing (Fig. 9E,F). The less severe phenotype is in line with the finding that K5-Cre-mediated deletion of ectodermal Cx26 between E12.5 and 15.5 does not occur in the entire ectoderm (cf. Fig. 7B–D; supplementary material Fig. S6). During embryogenesis K5–Cre activity leads to a mosaic deletion of Cx26 in the ectoderm. We conclude that a partial deletion of Cx26 in the ectoderm (under control of the K5 promoter) leads to malformation of peripheral lymphatic vessels, which results in lymphedema and embryonic death in utero.

Fig. 6.
Fig. 6.

Reduced Cx26 expression in Cx26+/Cx32 embryos and placentas. (A) Cx26 immunoblot analysis of E14.5 skin lysates: the Cx26+/+ and Cx26+/floxlacZ[Cx32] genotypes have similar amounts of Cx26 but in Cx26+/Cx32 lysates Cx26 is strongly reduced. (B) Quantification of Cx26 protein ratios in Cx26+/+, Cx26+/floxlacZ[Cx32] and Cx26+/Cx32 embryonic skin. Cx26 is significantly lower in Cx26+/Cx32 compared with Cx26+/floxlacZ[Cx32] and Cx26+/+ genotypes. (C) Cx26 immunoblot analysis of E12.5 placenta lysates: compared with the Cx26+/+ genotype, Cx26+/floxlacZ[Cx32] and Cx26+/floxlacZ[Cx32]neo have lower levels of Cx26 and the genotypes Cx26+/Cx32 and Cx26+/Cx32neo exhibit a further decrease in Cx26. (D) Quantification of Cx26 protein ratios in Cx26+/+, Cx26+/floxlacZ[Cx32] and Cx26+/Cx32 placentas. The Cx26 expression level is significantly decreased between Cx26+/+ and Cx26+/floxlacZ[Cx32] and between Cx26+/floxlacZ[Cx32] and Cx26+/Cx32. (E) Northern blot analysis of placental tissue showing Cx26 and Cx32 transcript levels of each genotype (E15.5). The genotypes Cx26+/floxlacZ[Cx32], Cx26+/floxlacZ[Cx32]neo, Cx26+/Cx32 and Cx26+/Cx32neo show similar levels of Cx26 mRNA (2.4 kb) but these are lower than in Cx26+/+. No endogenous Cx32 mRNA was detected but as expected Cx26+/Cx32 and Cx26+/Cx32neo placentas expressed transgenic Cx32 mRNA (2.5 kb). β-Actin was used as loading control in A and C. Cx26 protein levels in B and D were quantified by densitometric immunoblotting analysis using at least three embryos or placentas per genotype. Cx26 protein levels were normalized to β-actin expression as internal standard. Expression levels of Cx26+/+ embryonic skin or placenta were set to 1.0. Error bars indicate standard deviation (**P<0.01, ***P<0.001; independent Student's t-test).

Fig. 7.
Fig. 7.

Replacement of Cx26 with Cx32 in ectoderm leads to disturbed dermal lymphangiogenesis. (A) Appearance of Cx26+/floxlacZ[Cx32]:K5-Cre embryo on E17.5. The embryo exhibits severe lymphedemas and lacks the mandible (arrow), similar to Cx26+/Cx32 embryos. (B) Staining for β-gal in whole-mount Cx26+/floxlacZ[Cx32]:K5-Cre embryo and control littermate (Cx26+/floxlacZ[Cx32]) on E15.5. Because of K5–Cre-mediated deletion of the floxed lacZ reporter gene, β-gal activity in the epidermis of the Cx26+/floxlacZ[Cx32]:K5-Cre embryo is limited to Cx26-promoter active cells, which have no K5 promoter activity (cf. supplementary material Fig. S6). (C,D) Magnification of boxed regions of the embryos shown in B. (E-H) Whole-mount immunohistochemical staining of lymphatic capillaries on E14.5 using LYVE-1 as a marker: in comparison with the control littermate (Cx26+/+), the dermal lymphatic capillaries are severely reduced in the Cx26+/floxlacZ[Cx32]:K5-Cre embryo, showing only very few, mispatterned dermal lymphatics. Scale bars: 60 μm (A); 125 μm (B); 1 mm (C,D); 500 μm (E,F); 250 μm (G,H).

Fig. 8.
Fig. 8.

Relative transcript levels of factors relevant for lymphangiogenesis. Ligands, receptors and inhibitors were examined by qRT-PCR in E14.5 skin of Cx26+/floxlacZ[Cx32]:K5–Cre embryos and control littermates (Cx26+/+, Cx26+/floxlacZ[Cx32]). Error bars indicate standard deviation (*P<0.05, **P<0.01; independent Student's t-test).

Discussion

Our study provides the first demonstration that expression of Cx26 in ectodermal cells has an impact on the formation of a functional dermal lymphatic capillary system in mice. Both, ablation of ectodermal Cx26 or replacement of ectodermal Cx26 with Cx32 affect peripheral lymphangiogenesis and cause severe lymphedemas that result in embryonic death.

Because the fluid balance in Cx26+/Cx32 mutant embryos was altered after E14.5, we investigated the developing vascular systems. We found no evidence for cardiac or blood vascular defects. Therefore, we focused on embryonic lymphangiogenesis, which starts relatively late, i.e. subsequently to cardiovascular development. Although, the very first anlagen of the lymphatic vascular system, the jugular lymph sacs and also deep lymphatics, appeared normally developed in Cx26+/Cx32 embryos, the dermal lymphatic capillary system was nearly absent on E14.5. The reduction of peripheral dermal lymphatics in Cx26+/Cx32, Cx26+/floxlacZ[Cx32]:K5–Cre and Cx26fl/fl:K5–Cre embryos was detected at a developmental stage when the blood vascular system appeared normal. Thus we conclude that ablation or replacement of Cx26 with Cx32 in ectoderm specifically affected peripheral lymphangiogenesis. Hemorrhages observed on E17.5 were most probably caused by the large accumulation of lymph, leading to tissue compression in the embryonic body, but were not due to blood vessel defects. Taken together, the reduced dermal lymphatic vessel system of these mutants seems to be pivotal for defective tissue fluid homeostasis.

Because the skin appeared thinner in Cx26+/Cx32 embryos on E16.5, we cannot exclude functional abnormalities caused by the lack of Cx26 or the presence of Cx32. However, the epidermis appeared defective only in dorsal regions of later staged embryos (E16–18.5), when lymphedemas were very prominent there, but appeared normal in caudal regions. Thus we suppose that epidermal thinning is the consequence of mechanical stress due to large accumulation of lymph. Moreover, it seems unlikely that lymphedemas result from skin barrier defects because skin barrier formation, which is indicated by a differentiated cornified epidermis, does not start before E16.5–17.5 in mice (Hardman et al., 1998), and the first malformations and the lack of peripheral lymphatics in mutant embryos were observed 2–3 days earlier during development. Similar to the hemorrhages observed in late phase embryos, the epidermal abnormalities also seem to be a secondary effect that results from severe tissue swelling.

Fig. 9.
Fig. 9.

Peripheral lymphangiogenesis depends on ectodermal Cx26. (A,B) Cx26fl/fl:K5-Cre embryo in comparison with control littermate (Cx26+/fl) on E17.5. Cx26fl/fl:K5–Cre embryos are non viable and develop lymphedemas similar to Cx26+/Cx32 and Cx26+/floxlacZ[Cx32]:K5–Cre embryos. (C–F) Whole-mount immunohistochemical staining of LYVE-1 to show the lymphatic vasculature in Cx26fl/fl:K5-Cre and control (Cx26+/fl) embryos on E14.5. In comparison with the control littermate, the Cx26fl/fl:K5-Cre embryo lacks the fine dermal lymphatic vessels and shows coarsely connected dermal lymphatic capillaries. Scale bars: 1 mm (A–D); 250 μm (E,F).

How can the disturbed lymphangiogenesis in the ectodermal Cx26-deficient or Cx26KICx32 embryos be explained? Influence of the ectoderm on lymphangiogenesis has not been reported. Previously, it has been shown that blocking of VEGF-C and/or VEGF-D by ectopic expression of soluble VEGFR-3 in embryonic epidermis, controlled by the K14 promoter, leads to regression of dermal lymphatics (Mäkinen et al., 2001). Ectodermal cells might regulate lymphangiogenic factors that direct development of dermal lymphatics. Deletion of ectodermal Cx26 or replacement with Cx32 might change this function, thereby influencing peripheral lymphangiogenesis. However, instead of a decrease of VEGF-C, as possible mediator of lymphangiogenesis, we found an upregulation of this crucial factor in embryonic skin. In addition, its antagonist sVEGFR-2, and the transcripts of its receptors VEGFR-2 and VEGFR-3 were significantly downregulated. Therefore, it remains unclear whether upregulation of a ligand is able to compensate for the downregulation of its corresponding receptors or whether this imbalance contributes to the phenotype. In our Cx26+/floxlacZ[Cx32]:K5-Cre mouse line the upregulation of VEGF-C might represent a compensating mechanism for disturbed peripheral lymphangiogenesis induced by alteration of ectodermal Cx26 expression. Nevertheless, there is an imbalance of crucial factors known to contribute to the patterning of developing lymphatics.

Analyses of the mouse mutants Cx26+/floxlacZ[Cx32]:Prx1-Cre and Cx26+/floxlacZ[Cx32]:Tie2-Cre showed that neither the mesenchyme nor the blood endothelium are responsible for the lymphatic abnormalities in Cx26KICx32 mice. Instead, K5-Cre-mediated replacement of Cx26 with Cx32 (Cx26+/floxlacZ[Cx32]:K5-Cre embryos) or deletion of both Cx26 alleles resulted in similar phenotypic abnormalities as found in Cx26+/Cx32 embryos. Cx26+/floxlacZ[Cx32]:K5-Cre embryos, in which the replacement of Cx26 with Cx32 is restricted to the ectoderm, are embryonic lethal as a result of the lack of a peripheral lymphatic network. K5-Cre-mediated replacement of Cx26 with Cx32 exclusively occurred in ectodermal cells of the developing skin. Although the exchange did not take place in the whole ectoderm, the heterozygous replacement of Cx26 with Cx32 led to strong phenotypic abnormalities.

In case of Cx26+/Cx32 embryos, the expression of Cx26 protein was severely reduced when Cx32 protein was expressed ectopically. Previously, we had found that Cx26 protein was reduced in Cx32-deficient mice, although Cx26 mRNA was not affected (Nelles et al., 1996). However, consistent with the results of Nelles et al., no alteration in the level of Cx26 mRNA could be detected in Cx26+/Cx32 embryonic tissue. These findings indicate a transdominant effect of ectopic Cx32 expression on Cx26 expression, suggesting a posttranscriptional mechanism.

Loss of ectodermal Cx26 in Cx26fl/fl:K5-Cre embryos also induced lymphedemas and embryonic death as a result of regression of peripheral lymphatics. These findings demonstrate an essential role of ectodermal Cx26, because loss of Cx26 function produced malformed dermal lymphatics, which were not capable of maintaining tissue fluid homeostasis. But, because the lymphatic phenotype of some Cx26fl/fl:K5-Cre embryos was milder than in Cx26+/Cx32 or Cx26+/floxlacZ[Cx32]:K5-Cre embryos, but still embryonically lethal, it is possible that ectopic Cx32 expression aggravates the phenotypic abnormalities caused by reduction of Cx26. Stoichiometry of Cx26, Cx32 and heteromeric Cx26–Cx32 channels varies and enables cells to dynamically regulate their intercellular communication (Ayad et al., 2006). For instance heterotypic Cx26–Cx32 channels are selectively permeable to second messengers, such as cyclic nucleotides and inositol phosphates, whereas homomeric Cx26 or Cx32 channels are not (Ayad et al., 2006; Locke et al., 2004). We showed that the Cx26 protein level was significantly reduced in Cx26+/Cx32 embryos, but the remaining Cx26 still colocalized with ectopic Cx32. Probably, heteromeric Cx26–Cx32 or ectopic Cx32 channel properties additionally perturb Cx26-mediated intercellular signaling, which is needed for development of peripheral lymphatics.

Interestingly, in contrast to the ectoderm it is probable that in the placenta Cx32 can replace Cx26. It has been shown that deletion of Cx26 in the placenta leads to embryonic death on E10.5 (Gabriel et al., 1998). However, the observed reduction of Cx26 (~90%) in the placental labyrinth seems to be compensated by ectopic expression of Cx32, because these embryos are not growth retarded. Moreover this was shown with Cx26+/floxlacZ[Cx32]:K5–Cre and Cx26flox/flox:K5-Cre embryos, which exhibited similar phenotypic abnormalities as Cx26+/Cx32 embryos, although K5–Cre is not active in the placental labyrinth. These findings indicate that Cx32 can replace Cx26 in the placenta.

Our results suggest an essential role of Cx26-expressing ectodermal cells that can influence peripheral lymphangiogenesis, potentially by regulating lymphangiogenic signals from the early epidermis to the developing dermal lymphatic endothelium. Incidentally, it has been shown in humans that Cx26 expression is associated with lymphatic vessel invasion, large tumor size and poor prognosis in human breast cancer (Naoi et al., 2007). The authors conclude that Cx26 seems to enhance metastasis, probably by promoting invasion of the lymphatic vessels. Very recently, mutations in GJC2 (Cx47), another member of the connexin family, have been reported to cause primary lymphedema in humans (Ferrell et al., 2010; Ostergaard et al., 2011). With this in mind, it appears promising to further analyze connexin protein expression and function during development and maintenance of the lymphatic vessel system in mice and humans.

Materials and Methods

All mice used in this study were kept under standard housing conditions with a 12 hour:12 hour dark:light cycle and with food and water ad libitum. All experiments were carried out in accordance with local and state regulations for research with animals.

Cloning of the conditional Cx26KIlacZKICx32 targeting vector and ES cell transfection

The 1.9 kb 5′ and the 5.0 kb 3′ homologous regions were PCR amplified and cloned in pBluescript (Stratagene). A PCR fragment (106 bp) including 87 bp of the Cx26 (Gjb2) splice acceptor site and 6 bp of exon2 of Cx26 was cloned with SmaI in pBluescript and used twice in the final targeting vector, once upstream of the nls-lacZ and the second time upstream of the coding region of Cx32 (Gjb1). The Cx26 splice acceptor nls-lacZ reporter gene fragment, including a SV40 poly(A) signal, was blunt-end cloned (SmaI-SalI; Klenow) between the two loxP sites of the vector pHWlox1, which had been EcoRI digested and blunted. The Cx32 coding region (849 bp) was amplified by PCR (894 bp) introducing a NsiI restriction site 3′ of the stop codon and this fragment was cloned downstream of the Cx26 splice acceptor sequence. The introduced NsiI site was used to incorporate a neomycin selection cassette flanked by frt sites. The final construct was transformed in bacteria constitutively expressing Flp recombinase or Cre recombinase, to test for the deletion of the frt- or loxP-flanked regions, and was finally partially sequenced by the AGOWA company (Berlin, Germany) (Theis et al., 2000).

NotI-linearized targeting vector DNA (333 μg) were used to transfect HM1 (Magin et al., 1992) embryonic stem (ES) cells. Selection, PCR screening, Southern blot screening and blastocyst injections were performed as described previously (Buchholz et al., 1996) using an external 5′ Cx26 XbaI DNA fragment (1 kb) and an internal lacZ ClaI-NcoI DNA fragment (868 bp) as probes for Southern blot hybridization (DNA digested with EcoRI).

Generation and genotyping of Cx26+/floxlacZ[Cx32]neo, Cx26+/floxlacZ[Cx32], Cx26+/Cx32neo and Cx26+/Cx32 transgenic mice

By transfection with the targeting vector, one out of 680 ES cell clones had undergone homologous recombination and was used for blastocyst injection. After ES-cell transfection and germline transmission of the recombined allele, Cx26+/floxlacZ[Cx32]neo mice were obtained. These mice were bred with deleter-Flp mice (Rodriguez et al., 2000) to obtain Cx26+/floxlacZ[Cx32] mice, in which the frt-flanked selection cassette was removed. In order to obtain the embryonic-lethal genotypes Cx26+/Cx32neo or Cx26+/Cx32, ubiquitous Cre-recombinase-expressing mice (PGK–Cre) (Lallemand et al., 1998) were mated with Cx26+/floxlacZ[Cx32]neo or Cx26+/floxlacZ[Cx32] mice and the embryos were examined at various stages after timed matings. For cell-type-specific deletion of the loxP-flanked lacZ reporter gene, we bred female Cx26+/floxlacZ[Cx32] mice with male Tie2–Cre mice (Constien et al., 2001), male keratin 5 (K5)–Cre mice (Ramirez et al., 2004) or male Prx1–Cre mice (Logan et al., 2002). Cx26-mutated mice and embryos were genotyped (using tail DNA and yolk sac or tail DNA, respectively) by PCR using the primers Cx26 forward (5′-AGC AGC CTA GAG GAG GCT GTG TGT GC-3′), Cx26 reverse (5′-GGA CCC TTC GAT ACG GAC CTT CTG GG-3′), lacZ reverse (5′-CCA TTC AGG CTG CGC AAC TGT TGG-3′) and Cx32 reverse (5′-GCT CTC AGC AGC CAC CAC CAG CAC-3′).

Southern blot hybridization

Embryos (E15.5) of resulting transgenic genotypes were tested by Southern blot hybridization. For this purpose, isolated DNA of each genotype was digested with BglII. After electrophoretic separation on agarose gels, blotting onto Hybond N+ membranes (Amersham Biosciences) and ultraviolet-mediated crosslinking for fixation, hybridization was performed under stringent conditions using the Quick Hyb solution (Stratagene) at 68°C for 1.5 hours. An ~1 kb XbaI fragment of the 5′ homology region of Cx26 was used as the external probe and a 0.5 kb NcoI-XbaI fragment of the Cx32 coding region served as the internal probe.

Northern blot hybridization

Total RNA from placentas (E12.5) was collected using TRIzol™ (Invitrogen) according to the manufacturer's protocol. Loaded aliquots were electrophoretically separated, blotted and hybridized as described previously (Hennemann et el., 1992). For detection of Cx26 mRNA, a 460 bp BglII-BstXI fragment of the Cx26 coding region, and for Cx32 mRNA, a 450 bp EcoNI fragment of the Cx32 coding region were used for generation of the probes.

RNA preparation and cDNA synthesis

The skin of Cx26+/floxlacZ[Cx32]:K5–Cre embryos (n=4) and Cx26+/+ and Cx26+/floxlacZ[Cx32] control littermates (n=3) was dissected and pooled on E14.5. Total RNA from embryonic skin was collected using TRIzol™ (Invitrogen) according to the manufacturer's protocol. Reverse transcription of 2 μg RNA included a DNase I (Invitrogen) treatment and was performed using Oligo(dT)18 primer (Invitrogen) and Moloney murine leukemia virus reverse transcriptase (Invitrogen) according to the protocol supplied.

Quantitative PCR

Quantitative real-time PCR was performed in triplicate using an ABI Prism 7300 Sequence Detector (Applied Biosystems). For qRT-PCR, 1 μl of first strand cDNA solution was diluted in a final volume of 20 μl containing 0.2 pmol of each gene-specific primer and 10 μl Power SYBR® Green PCR Master Mix (Applied Biosystems) including a ROX™ dye as an internal passive reference. The signal intensities of VEGF-C, VEGF-D, VEGFR-2, sVEGFR-2, VEGFR-3 and neuropilin-2 transcripts were quantified using serial dilutions of gene-specific standard cDNA. β-Actin cDNA was used as a standard for normalization. Subsequent melting curve analysis was used to determine the specificity of the PCR fragments. A tenfold dilution series of purified PCR products ranging from 1 pg to 0.1 fg were used to create a standard curve for each gene. The PCR reactions were carried out in triplicate. All primer sequences are shown in supplementary material Table S2.

Immunoblot analysis

Embryos or placentas were solubilized in RIPA buffer or 1′ Complete buffer (Roche) and sonicated three times for 10 seconds while being incubated on ice. Immunoblotting was performed as described by Dobrowolski et al. using mouse anti-Cx26 (1:750; Zymed) and goat anti-mouse horseradish-peroxidase (HRP)-conjugated antibodies (1:10,000; Dianova, Hamburg, Germany) (Dobrowolski et al., 2008). The concentration of the loaded proteins was analyzed by immunoblotting using anti-β-actin antibodies (1:500; Sigma). Protein expression was quantified using the Herolab E.A.S.Y. Win32 quantification software (Herolab GmbH, Wiesloch, Germany).

Histological analysis of embryos

Embryos were fixed in 4% paraformaldehyde (PFA) or Bouin solution, dehydrated in a graded series of ethanol, embedded in Paraplast plus (Sherwood Medical Industries, St Louis, MO). Paraffin serial sections (4–7 μm) were stained with Hematoxylin and acid fuchsin or Hematoxylin and Eosin.

Whole-mount staining of embryos for β-gal

After dissection, E10.5–14.5 embryos were fixed for 30–120 minutes in 4% PFA or 0.2% glutaraldehyde or both at 4°C. Staining for β-gal was performed as previously described (Krüger et al., 2000).

Staining of cryosections for β-gal followed by immunohistochemical analysis

Cryostat sections (10–25 μm) were fixed with 0.2% glutaraldehyde for 5 minutes, or for combined β-gal staining and immunohistochemical analysis on the same section, in 2% PFA and 0.1% glutaraldehyde for 2 minutes. After washing steps, staining for β-gal was performed using the substrate 5-bromo-4-chloro-3-indolyl-β-galactoside (X-gal). After X-gal staining, sections were post-fixed for 5 minutes in 4% PFA and immunohistochemical staining was carried out using the M.O.M.™ Immunodetection Kit (Vector Laboratories) following the manufacturer's protocol. Sections were incubated with rabbit anti-LYVE-1 IgG (1:200; Reliatech, Braunschweig, Germany). Secondary antibody incubation was then carried out with biotin-conjugated goat anti-rabbit IgG (1:500; Zymed) for 45 minutes at room temperature. The immunocomplexes formed were visualized by peroxidase enzyme reaction containing 3.3′-diaminobenzidine tetrahydrochloride (Sigma-Aldrich, Steinheim, Germany) and <3 mmol/l H2O2 for ~10 min. After staining, sections were rapidly washed, counter-stained with 0.1% Eosin and mounted in Entellan (Merck, Hohenbrunn, Germany).

Immunofluorescence analysis

Embryos and placentas were frozen on dry ice at –70°C and embedded in TissueTec (Miles Inc., Elkhart, USA). Cryostat sections (12–25 μm) were fixed for 5 minutes with absolute ethanol at –20°C or with 4% PFA, washed three times with PBS without calcium and magnesium (PBS−) and blocked in 4% normal goat serum or 4% bovine serum albumin, 0.1% Triton X-100 in PBS− for 1 hour. The following primary antibodies were used: rabbit anti-Cx26 (1:400–1:800; Zymed), mouse anti-Cx32 (1:300–1:600; Zymed) and rabbit anti-Prox1 (1:100; Reliatech) in blocking solution. After incubation overnight, the sections were washed three times in PBS− and afterwards incubated for 1 hour at room temperature with secondary antibodies: goat anti-rabbit conjugated to Alexa Fluor 594 (1:1000) and goat anti-mouse conjugated to Alexa Fluor 488 (1:1000; Mobitech, Goettingen, Germany). Nuclei were stained with Hoechst 33258 (1:20,000; Sigma-Aldrich). After additional PBS− washing steps, sections were embedded in Dako Glycergel mounting medium (Dako, Hamburg, Germany) and analyzed by confocal microscopy (LSM 510 and LSM 710; Zeiss).

Whole-mount immunohistochemical analysis

Immunohistochemical staining of whole-mount embryos was performed as described previously (Nagy et al., 2003) but modified by using the M.O.M.™ Immunodetection Kit (Vector Laboratories) following the manufacturer's protocol. As primary antibodies, rabbit anti-LYVE-1 IgG and rat anti-MECA-32 IgG (1:50, BD Pharmigen, Frankling Lakes, USA) were used.

Acknowledgments

We thank Gerda Hertig and Christine Siegmund for excellent technical assistance. This work is part of the PhD thesis of N.D. who was generously supported by a stipend from the Jürgen-Manchot Foundation (Düsseldorf). The Cx26+/floxlacZ[Cx32]neo mice were generated during the PhD work of N.P. In addition, K.W. acknowledges support of the German Research Foundation (SFB 645, B2) and Bonn University.

Footnotes

  • Accepted April 4, 2011.

References

  1. ↵
    1. Albuquerque, R. J.,
    2. Hayashi, T.,
    3. Cho, W. G.,
    4. Kleinman, M. E.,
    5. Dridi, S.,
    6. Takeda, A.,
    7. Baffi, J. Z.,
    8. Yamada, K.,
    9. Kaneko, H.,
    10. Green, M. G.,
    11. et al
    . (2009). Alternatively spliced vascular endothelial growth factor receptor-2 is an essential endogenous inhibitor of lymphatic vessel growth. Nat. Med. 15, 1023-1030.
    OpenUrlCrossRefPubMedWeb of Science
  2. ↵
    1. Ayad, W. A.,
    2. Locke, D.,
    3. Koreen, I. V. and
    4. Harris, A. L.
    (2006). Heteromeric, but not homomeric, connexin channels are selectively permeable to inositol phosphates. J. Biol. Chem. 281, 16727-16739.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Banerji, S.,
    2. Ni, J.,
    3. Wang, S. X.,
    4. Clasper, S.,
    5. Su, J.,
    6. Tammi, R.,
    7. Jones, M. and
    8. Jackson, D. G.
    (1999). LYVE-1, a new homologue of the CD44 glycoprotein, is a lymph-specific receptor for hyaluronan. J. Cell Biol. 144, 789-801.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Buchholz, F.,
    2. Angrand, P. O. and
    3. Stewart, A. F.
    (1996). A simple assay to determine the functionality of Cre or FLP recombination targets in genomic manipulation constructs. Nucleic Acids Res. 24, 3118-3119.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Buttler, K.,
    2. Kreysing, A.,
    3. von Kaisenberg, C. S.,
    4. Schweigerer, L.,
    5. Gale, N.,
    6. Papoutsi, M. and
    7. Wilting, J.
    (2006). Mesenchymal cells with leukocyte and lymphendothelial characteristics in murine embryos. Dev. Dyn. 235, 1554-1562.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Buttler, K.,
    2. Ezaki, T. and
    3. Wilting, J.
    (2008). Proliferating mesodermal cells in murine embryos exhibiting macrophage and lymphendothelial characteristics. BMC Dev. Biol. 8, 43.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Cohen-Salmon, M.,
    2. Ott, T.,
    3. Michel, V.,
    4. Hardelin, J. P.,
    5. Perfettini, I.,
    6. Eybalin, M.,
    7. Wu, T.,
    8. Marcus, D. C.,
    9. Wangemann, P.,
    10. Willecke, K.,
    11. et al
    . (2002). Targeted ablation of connexin26 in the inner ear epithelial gap junction network causes hearing impairment and cell death. Curr. Biol. 12, 1106-1111.
    OpenUrlCrossRefPubMedWeb of Science
  8. ↵
    1. Constien, R.,
    2. Forde, A.,
    3. Liliensiek, B.,
    4. Grone, H. J.,
    5. Nawroth, P.,
    6. Hammerling, G. and
    7. Arnold, B.
    (2001). Characterization of a novel EGFP reporter mouse to monitor Cre recombination as demonstrated by a Tie2 Cre mouse line. Genesis 30, 36-44.
    OpenUrlCrossRefPubMedWeb of Science
  9. ↵
    1. Dobrowolski, R.,
    2. Sasse, P.,
    3. Schrickel, J. W.,
    4. Watkins, M.,
    5. Kim, J. S.,
    6. Rackauskas, M.,
    7. Troatz, C.,
    8. Ghanem, A.,
    9. Tiemann, K.,
    10. Degen, J.,
    11. et al
    . (2008). The conditional connexin43G138R mouse mutant represents a new model of hereditary oculodentodigital dysplasia in humans. Hum. Mol. Genet. 17, 539-554.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Dobrowolski, R.,
    2. Hertig, G.,
    3. Lechner, H.,
    4. Wörsdörfer, P.,
    5. Wulf, V.,
    6. Dicke, N.,
    7. Eckert, D.,
    8. Bauer, R.,
    9. Schorle, H. and
    10. Willecke, K.
    (2009). Loss of connexin43-mediated gap junctional coupling in the mesenchyme of limb buds leads to altered expression of morphogensin mice. Hum. Mol. Genet. 18, 2899-2911.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Ferrell, R. E.,
    2. Baty, C. J.,
    3. Kimak, M. A.,
    4. Karlsson, J. M.,
    5. Lawrence, E. C.,
    6. Franke-Snyder, M.,
    7. Meriney, S. D.,
    8. Feingold, E. and
    9. Finegold, D. N.
    (2010). GJC2 missense mutations cause human lymphedema. Am. J. Hum. Genet. 86, 943-948.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Frank, M.,
    2. Eiberger, B.,
    3. Janssen-Bienhold, U.,
    4. de Sevilla Müller, L. P.,
    5. Tjarks, A.,
    6. Kim, J. S.,
    7. Maschke, S.,
    8. Dobrowolski, R.,
    9. Sasse, P.,
    10. Weiler, R.,
    11. et al
    . (2010). Neuronal connexin-36 can functionally replace connexin-45 in mouse retina but not in the developing heart. J. Cell Sci. 123, 3605-3615.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Gabriel, H. D.,
    2. Jung, D.,
    3. Bützler, C.,
    4. Temme, A.,
    5. Traub, O.,
    6. Winterhager, E. and
    7. Willecke, K.
    (1998). Transplacental uptake of glucose is decreased in embryonic lethal connexin26-deficient mice. J. Cell Biol. 140, 1453-1461.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Hardman, M. J.,
    2. Sisi, P.,
    3. Banbury, D. N. and
    4. Byrne, C.
    (1998). Patterned acquisition of skin barrier function during development. Development 125, 1541-1552.
    OpenUrlAbstract
  15. ↵
    1. Hennemann, H.,
    2. Schwarz, H. J. and
    3. Willecke, K.
    (1992). Characterization of gap junction genes expressed in F9 embryonic carcinoma cells: molecular cloning of mouse connexin31 and -45 cDNAs. Eur. J. Cell Biol. 57, 51-58.
    OpenUrlPubMedWeb of Science
  16. ↵
    1. Karkkainen, M. J.,
    2. Haiko, P.,
    3. Sainio, K.,
    4. Partanen, J.,
    5. Taipale, J.,
    6. Petrova, T. V.,
    7. Jeltsch, M.,
    8. Jackson, D. G.,
    9. Talikka, M.,
    10. Rauvala, H.,
    11. et al
    . (2004). Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat. Immunol. 5, 74-80.
    OpenUrlCrossRefPubMedWeb of Science
  17. ↵
    1. Karpanen, T. and
    2. Alitalo, K.
    (2008). Molecular biology and pathology of lymphangiogenesis. Annu. Rev. Pathol. 3, 367-397.
    OpenUrlCrossRefPubMedWeb of Science
  18. ↵
    1. Krüger, O.,
    2. Plum, A.,
    3. Kim, J. S.,
    4. Winterhager, E.,
    5. Maxeiner, S.,
    6. Hallas, G.,
    7. Kirchhoff, S.,
    8. Traub, O.,
    9. Lamers, W. H. and
    10. Willecke, K.
    (2000). Defective vascular development in connexin 45-deficient mice. Development 127, 4179-4193.
    OpenUrlAbstract
  19. ↵
    1. Lallemand, Y.,
    2. Luria, V.,
    3. Haffner-Krausz, R. and
    4. Lonai, P.
    (1998). Maternally expressed PGK-Cre transgene as a tool for early and uniform activation of the Cre site-specific recombinase. Transgenic Res. 7, 105-112.
    OpenUrlCrossRefPubMedWeb of Science
  20. ↵
    1. Locke, D.,
    2. Stein, T.,
    3. Davies, C.,
    4. Morris, J.,
    5. Harris, A. L.,
    6. Evans, W. H.,
    7. Monaghan, P. and
    8. Gusterson, B.
    (2004). Altered permeability and modulatory character of connexin channels during mammary gland development. Exp. Cell Res. 298, 643-660.
    OpenUrlCrossRefPubMedWeb of Science
  21. ↵
    1. Logan, M.,
    2. Martin, J. F.,
    3. Nagy, A.,
    4. Lobe, C.,
    5. Olson, E. N. and
    6. Tabin, C. J.
    (2002). Expression of Cre recombinase in the developing mouse limb bud driven by a Prxl enhancer. Genesis 33, 77-80.
    OpenUrlCrossRefPubMedWeb of Science
  22. ↵
    1. Magin, T. M.,
    2. McWhir, J. and
    3. Melton, D. W.
    (1992). A new mouse embryonic stem cell line with good germ line contribution and gene targeting frequency. Nucleic Acids Res. 20, 3795-3796.
    OpenUrlFREE Full Text
  23. ↵
    1. Mäkinen, T.,
    2. Jussila, L.,
    3. Veikkola, T.,
    4. Karpanen, T.,
    5. Kettunen, M. I.,
    6. Pulkkanen, K. J.,
    7. Kauppinen, R.,
    8. Jackson, D. G.,
    9. Kubo, H.,
    10. Nishikawa, S.,
    11. et al
    . (2001). Inhibition of lymphangiogenesis with resulting lymphedema in transgenic mice expressing soluble VEGF receptor-3. Nat. Med. 7, 199-205.
    OpenUrlCrossRefPubMedWeb of Science
  24. ↵
    1. Mäkinen, T.,
    2. Norrmen, C. and
    3. Petrova, T. V.
    (2007). Molecular mechanisms of lymphatic vascular development. Cell. Mol. Life Sci. 64, 1915-1929.
    OpenUrlCrossRefPubMedWeb of Science
  25. ↵
    1. Nagy, A.,
    2. Gertsenstein, M.,
    3. Vintersten, K. and
    4. Behringer, R.
    (eds) (2003). Immunhistochemistry of whole-mount embryos. In Manipulating the Mouse Embryo: A Laboratory Manual, 3rd edn, pp. 666-669. Cold Spring Harbor, NY: Cold Spring Harbor Press.
  26. ↵
    1. Naoi, Y.,
    2. Miyoshi, Y.,
    3. Taguchi, T.,
    4. Kim, S. J.,
    5. Arai, T.,
    6. Tamaki, Y. and
    7. Noguchi, S.
    (2007). Connexin26 expression is associated with lymphatic vessel invasion and poor prognosis in human breast cancer. Breast Cancer Res. Treat. 106, 11-17.
    OpenUrlCrossRefPubMedWeb of Science
  27. ↵
    1. Neijssen, J.,
    2. Herberts, C.,
    3. Drijfhout, J. W.,
    4. Reits, E.,
    5. Janssen, L. and
    6. Neefjes, J.
    (2005). Cross-presentation by intercellular peptide transfer through gap junctions. Nature 434, 83-88.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Nelles, E.,
    2. Bützler, C.,
    3. Jung, D.,
    4. Temme, A.,
    5. Gabriel, H. D.,
    6. Dahl, U.,
    7. Traub, O.,
    8. Stümpel, F.,
    9. Jungermann, K.,
    10. Zielasek, J.,
    11. et al
    . (1996). Defective propagation of signals generated by sympathetic nerve stimulation in the liver of connexin32-deficient mice. Proc. Natl. Acad. Sci. USA 93, 9565-9570.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Nicholson, B.,
    2. Dermietzel, R.,
    3. Teplow, D.,
    4. Traub, O.,
    5. Willecke, K. and
    6. Revel, J. P.
    (1987). Two homologous protein components of hepatic gap junctions. Nature 329, 732-734.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Ny, A.,
    2. Koch, M.,
    3. Schneider, M.,
    4. Neven, E.,
    5. Tong, R. T.,
    6. Maity, S.,
    7. Fischer, C.,
    8. Plaisance, S.,
    9. Lambrechts, D.,
    10. Heligon, C.,
    11. et al
    . (2005). A genetic Xenopus laevis tadpole model to study lymphangiogenesis. Nat. Med. 11, 998-1004.
    OpenUrlCrossRefPubMedWeb of Science
  31. ↵
    1. Oliver, G. and
    2. Srinivasan, R. S.
    (2010). Endothelial cell plasticity: how to become and remain a lymphatic endothelial cell. Development 137, 363-372.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Ostergaard, P.,
    2. Simpson, M. A.,
    3. Brice, G.,
    4. Mansour, S.,
    5. Connell, F. C.,
    6. Onoufriadis, A.,
    7. Child, A. H.,
    8. Hwang, J.,
    9. Kalidas, K.,
    10. Mortimer, P. S.,
    11. et al
    . (2011). Rapid identification of mutations in GJC2 in primary lymphoedema using whole exome sequencing combined with linkage analysis with delineation of the phenotype. J. Med. Genet. 48, 251-255.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Ramirez, A.,
    2. Page, A.,
    3. Gandarillas, A.,
    4. Zanet, J.,
    5. Pibre, S.,
    6. Vidal, M.,
    7. Tusell, L.,
    8. Genesca, A.,
    9. Whitaker, D. A.,
    10. Melton, D. W.,
    11. et al
    . (2004). A keratin K5Cre transgenic line appropriate for tissue-specific or generalized Cre-mediated recombination. Genesis 39, 52-57.
    OpenUrlCrossRefPubMedWeb of Science
  34. ↵
    1. Rodriguez, C. I.,
    2. Buchholz, F.,
    3. Galloway, J.,
    4. Sequerra, R.,
    5. Kasper, J.,
    6. Ayala, R.,
    7. Stewart, A. F. and
    8. Dymecki, S. M.
    (2000). High-efficiency deleter mice show that FLPe is an alternative to Cre-loxP. Nat. Genet. 25, 139-140.
    OpenUrlCrossRefPubMedWeb of Science
  35. ↵
    1. Sato, T. N.,
    2. Qin, Y.,
    3. Kozak, C. A. and
    4. Audus, K. L.
    (1993). Tie-1 and tie-2 define another class of putative receptor tyrosine kinase genes expressed in early embryonic vascular system. Proc. Natl. Acad. Sci. USA 90, 9355-9358.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Simon, A. M. and
    2. McWhorter, A. R.
    (2003). Decreased intercellular dye-transfer and downregulation of non-ablated connexins in aortic endothelium deficient in connexin37 or connexin40. J. Cell Sci. 116, 2223-2236.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. Söhl, G. and
    2. Willecke, K.
    (2003). An update on connexin genes and their nomenclature in mouse and man. Cell Commun. Adhes. 10, 173-180.
    OpenUrlCrossRefPubMedWeb of Science
  38. ↵
    1. Soriano, P.
    (1999). Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat. Genet. 21, 70-71.
    OpenUrlCrossRefPubMedWeb of Science
  39. ↵
    1. Srinivasan, R. S.,
    2. Dillard, M. E.,
    3. Lagutin, O. V.,
    4. Lin, F. J.,
    5. Tsai, S.,
    6. Tsai, M. J.,
    7. Samokhvalov, I. M. and
    8. Oliver, G.
    (2007). Lineage tracing demonstrates the venous origin of the mammalian lymphatic vasculature. Genes Dev. 21, 2422-2432.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Takakura, N.,
    2. Huang, X. L.,
    3. Naruse, T.,
    4. Hamaguchi, I.,
    5. Dumont, D. J.,
    6. Yancopoulos, G. D. and
    7. Suda, T.
    (1998). Critical role of the TIE2 endothelial cell receptor in the development of definitive hematopoiesis. Immunity 9, 677-686.
    OpenUrlCrossRefPubMedWeb of Science
  41. ↵
    1. Tammela, T. and
    2. Alitalo, K.
    (2010). Lymphangiogenesis: molecular mechanisms and future promise. Cell 140, 460-476.
    OpenUrlCrossRefPubMedWeb of Science
  42. ↵
    1. Tammela, T.,
    2. Petrova, T. V. and
    3. Alitalo, K.
    (2005). Molecular lymphangiogenesis: new players. Trends Cell Biol. 15, 434-441.
    OpenUrlCrossRefPubMedWeb of Science
  43. ↵
    1. Theis, M.,
    2. Magin, T. M.,
    3. Plum, A. and
    4. Willecke, K.
    (2000). General or cell type-specific deletion and replacement of connexin-coding DNA in the mouse. Methods 20, 205-218.
    OpenUrlCrossRefPubMedWeb of Science
  44. ↵
    1. Wigle, J. T. and
    2. Oliver, G.
    (1999). Prox1 function is required for the development of the murine lymphatic system. Cell 98, 769-778.
    OpenUrlCrossRefPubMedWeb of Science
  45. ↵
    1. Wigle, J. T.,
    2. Harvey, N.,
    3. Detmar, M.,
    4. Lagutina, I.,
    5. Grosveld, G.,
    6. Gunn, M. D.,
    7. Jackson, D. G. and
    8. Oliver, G.
    (2002). An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. EMBO J. 21, 1505-1513.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    1. Wilting, J.,
    2. Aref, Y.,
    3. Huang, R.,
    4. Tomarev, S. I.,
    5. Schweigerer, L.,
    6. Christ, B.,
    7. Valasek, P. and
    8. Papoutsi, M.
    (2006). Dual origin of avian lymphatics. Dev. Biol. 292, 165-173.
    OpenUrlCrossRefPubMedWeb of Science
  47. ↵
    1. Xu, Y.,
    2. Yuan, L.,
    3. Mak, J.,
    4. Pardanaud, L.,
    5. Caunt, M.,
    6. Kasman, I.,
    7. Larrivee, B.,
    8. Del, T. R.,
    9. Suchting, S.,
    10. Medvinsky, A.,
    11. et al
    . (2010). Neuropilin-2 mediates VEGF-C-induced lymphatic sprouting together with VEGFR3. J. Cell Biol. 188, 115-130.
    OpenUrlAbstract/FREE Full Text
Back to top

 Download PDF

Email
Research Article
Peripheral lymphangiogenesis in mice depends on ectodermal connexin-26 (Gjb2)
Nikolai Dicke, Nicole Pielensticker, Joachim Degen, Julia Hecker, Oliver Tress, Tobias Bald, Alexandra Gellhaus, Elke Winterhager, Klaus Willecke
Journal of Cell Science 2011 124: 2806-2815; doi: 10.1242/jcs.084186
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo
Citation Tools
Alerts

Article navigation

Other journals from The Company of Biologists

Advertisement

Introducing FocalPlane’s new Community Manager, Esperanza Agullo-Pascual

We are pleased to welcome Esperanza to the Journal of Cell Science team. The new Community Manager for FocalPlane, Esperanza is joining us from the Microscopy Core at Mount Sinai School of Medicine. Find out more about Esperanza in her introductory post over on FocalPlane.

New funding scheme supports sustainable events

As part of our Sustainable Conferencing Initiative, we are pleased to announce funding for organisers that seek to reduce the environmental footprint of their event. The next deadline to apply for a Scientific Meeting grant is 26 March 2021.

Read & Publish participation continues to grow

"Alongside pre-printing for early documentation of work, such mechanisms are particularly helpful for early-career researchers like me.”

Dr Chris MacDonald (University of York) shares his experience of publishing Open Access as part of our growing Read & Publish initiative. We now have over 150 institutions in 15 countries and four library consortia taking part – find out more and view our full list of participating institutions.

Cell scientist to watch: Romain Levayer

In an interview, Romain Levayer talks about starting his own lab, his love for preprints and his experience of balancing parenting with his research goals.

Live lactating mammary tissue

In a stunning video, Stewart et al. demonstrate warping of the alveolar unit due to basal cell-generated force as part of their recent work investigating roles for mechanically activated ion channels in lactation and involution.

Visit our YouTube channel to watch more videos from JCS, our sister journals and the Company.

JCS and COVID-19

For more information on measures Journal of Cell Science is taking to support the community during the COVID-19 pandemic, please see here.

If you have any questions or concerns, please do not hestiate to contact the Editorial Office.