The Tight Skin Mouse: Demonstration of Mutant Fibrillin-1 Production and Assembly into Abnormal Microfibrils

Bacterial collagenase (type 1A), PMSF, N-ethylmaleimide (NEM), benzamidine, guanidinium chloride (GuHCl), protein A–Sepharose colloidal gold conjugate, protein A–Sepharose, and Triton X-100 were obtained from Sigma Chemical Co. (Poole, Dorset, UK). Sepharose CL-2B was supplied by Pharmacia-LKB (Milton Keynes, Bucks, UK). MultiMark™ SDS-PAGE molecular weight markers were supplied by Novex (San Diego, CA). Normal donkey serum and Texas red were purchased from Jackson ImmunoResearch (West Grove, PA), and Mowiol was purchased from Hoechst, (Frankfurt-Hoechst, Germany). ³⁵[S]cys/met Translabel and ⁴⁵CaCl₂ was obtained from ICN Biomedicals, Inc., (Meckenheim, Germany). Chelex 100 was obtained from (Bio-Rad Laboratories, Hercules, CA). The antifibrillin antisera used were a polyclonal antibody (9022) raised to fibrillin-1 PF2 fragment (20) that was kindly provided by R. Glanville (Shriner's Hospital, Portland, OR) and a polyclonal antibody (5077) raised to intact microfibrils (17, 20, 27).

Mice of the C57BL/6-pa +/+Tsk genotype were originally purchased from Jackson Laboratory (Bar Harbor, ME). The Tsk mutation was transferred onto the C57BL/6J (B6) background by sequential backcrossing at the Jefferson Medical College (Philadelphia, PA). The B6-Tsk/+ and B6-+/+ mice used for these studies were at the N6 generation. The mutant Tsk phenotype was determined by manual assessment of the thickness and tightness of the skin in the intercapsular region as well as by Southern blot analyses to distinguish the presence or absence of the duplicated region of the Fbn 1 gene (37).

Dermal fibroblasts were established from normal and Tsk/+ mice and were seeded at a density of 5 × 10⁵ per 35-mm dish after passage. When confluent, the cultures were either labeled continuously in the presence of 100 μCi of ³⁵[S]cys/met per 0.5ml cys- and met-free minimal essential medium per well for 17 h, or pulse labeled with 100 μCi ³⁵[S]cys/met for 60 min, and then chased for 24 h. Fibrillin was radioimmunoprecipitated from the continuously labeled culture medium, which was preabsorbed with normal rabbit serum and protein A–Sepharose overnight at 4°C. After removal of Sepharose beads, the media were incubated with the PF2 antibody or normal rabbit serum (25 μl per 0.5-ml aliquot) for 19 h at 4°C. Immune complexes were adsorbed to a suspension of 50 μl of protein A–Sepharose for 2 h at room temperature and then washed three times in radioimmunoprecipitation assay buffer. The bound proteins were released by boiling in 75 μl of double-strength sample buffer containing 10% β-mercaptoethanol. Proteins were separated on 4% SDS-PAGE gels. Aliquots of crude medium were added to equal volumes of double-strength sample buffer and then analyzed on the same gel (27).

Cryostat sections (7 μm) of normal and Tsk/+ mice skin were fixed in acetone for 10 min. Nonspecific binding sites were blocked by incubation with 10% normal donkey serum/1% BSA in PBS for 30 min at room temperature. This step was followed by incubation in primary antibody (PF2 antibody or 5507; 1:100 in PBS) overnight at room temperature. Bound antibody was visualized using a 1:100 dilution of donkey Fab2 against rabbit IgG coupled to Texas red for 30 min at room temperature. Preparations were mounted in Mowiol in 0.01 M Tris-HCl, pH 8.6. Preparations were examined using an invert device (model LSM 410; Carl Zeiss, Inc., Oberkochen, Germany). A 40-fold oil immersion objective (numerical aperture 1.3) was used. Single optical sections (0.6–0.7 μm) were obtained with identical scanning settings and excitation conditions to distinguish differences in fluorescence intensities between Tsk/+ and normal mouse skin.

Skin biopsies were minced, fixed in 2.5% glutaraldehyde in 100 mM cacodylate, pH 7.2, for 48 h, osmicated en bloc, and then embedded in Epon 812 (Polysciences Inc., Warington, PA). Ultrathin sections were counterstained with uranyl acetate and lead citrate and then viewed with an electron microscope (model EM 401; Philips Electron Optics, Eindhoven, The Netherlands) at an accelerating voltage of 60 kV.

Triplicate preparations of intact native microfibrils were isolated from normal and Tsk/+ mouse skin using a well-defined methodology (17, 19, 36). Briefly, tissues were disrupted by bacterial collagenase digestion in 0.05 M Tris-HCl, pH 7.4, containing 0.4 M NaCl and 0.01 M CaCl₂ in the presence of proteinase inhibitors (5 mM PMSF, 2 mM NEM, 5 mM benzamidine). Microfibrils were isolated as high M_r assemblies in the void volume of a Sepharose CL-2B gel filtration chromatography column (15 × 1 cm; Pharmacia Biotech Ltd., Herts, UK). All microfibril preparations were examined directly by rotary shadowing–electron microscopy and scanning transmission electron microscopy (STEM). Microfibrils were incubated for various lengths of time at 20°C in 5 mM EDTA or 5 mM CaCl₂, or at 37°C in 4 M GuHCl.

A suspension of intact microfibrils was equilibrated in 0.025 M Tris-HCl, pH 7.4, containing 0.2 M NaCl with or without 5 mM EDTA and then incubated for 2 or 18 h at 37°C in the presence of trypsin (Sigma Chemical Co.) (1 μg:100 μl) or neutrophil elastase (gift from D. Woolley, University of Manchester, Manchester, UK) (100 ng:100 μl). Consequences for microfibril integrity were monitored by rotary shadowing.

Preparations of intact microfibrils were adsorbed directly onto carbon-coated grids (2–3-nm film thickness) and metal shadowed with a tungsten/ platinum filament at an angle of 4° (17, 36). Grids were visualized using an electron microscope (model 1200EX; JEOL USA, Inc., Peabody, MA) at an accelerating voltage of 100 kV.

STEM mass analysis is a well-established technique that has provided quantitative data on mass per unit length and axial mass distribution on unstained and unshadowed fibrillin-containing microfibrils and other macromolecules (11, 12, 24, 35, 36). Preparations of intact native microfibrils were analyzed by STEM in a JEOL 1200EX transmission electron microscope as previously described (36). Mean axial repeat measurements for each microfibril preparation were calculated from 40–50 axial repeat measurements across 10–30 periods per microfibril, with an error of ±0.34%. Five measurements of mass per unit length were taken per microfibril, with 40–50 microfibrils per grid examined. Axial mass distributions were averages involving 10 measurements per microfibril. In all cases, these STEM measurements were repeated on triplicate preparations. The axial repeat was the distance from the beginning of one bead to the beginning of the next (effectively one bead plus one interbead domain). Analysis of variance analysis of axial repeat, mass per unit length, and total mass per axial repeat data was performed.

Metabolic labeling of normal mouse dermal fibroblasts and subsequent SDS-PAGE under reducing conditions revealed a single protein band of ∼330 kD that comigrated with fibrillin from normal human fibroblasts (data not shown) and was identified by the PF2 antibody (Fig. 1,A). In contrast, the Tsk/+ dermal fibroblast culture medium contained an additional protein of ∼450 kD that was also recognized by these antibodies and was therefore identified as the gene product of the larger mutant Tsk Fbn 1 gene. Bands corresponding to normal and Tsk fibrillin-1 in the Tsk/+ fibroblast radioimmunoprecipitates were of comparable intensity. No specific protein bands were immunoprecipitated with normal rabbit serum. Pulse-chase analysis revealed that Tsk fibrillin-1 could be recovered from the extracellular matrix of the fibroblast cultures (Fig. 1 B).

A confocal laser scanning analysis of immunostaining of the cutaneous microfibrillar network appeared substantially more intense in the Tsk/+ skin than in the normal mouse with the PF2 antibody, and far more details of this structure were discernible (Fig. 2,A). In contrast, the antibody 5507 immunostained microfibrils in both normal and mutant mice, but the immunostaining intensity was lower in the Tsk/+ mouse under identical scanning conditions (Fig. 2 B). This difference was most pronounced with the microfibrils around hair follicles.

A transmission electron microscopy study revealed that, in comparison with normal mice, Tsk/+ mice skin had more prominent microfibrillar “clusters” in the upper dermis, although there was a noticable absence of well-packed elastic fibrils with very electron-dense cores (Fig. 3). Tsk/+ mouse microfibrils, with or without associated elastin, frequently appeared blurred without a discernible striated pattern. The elastin appeared less electron dense and less compact than those in samples from normal animals.

Abundant beaded microfibrils were isolated from littermate pairs of normal and Tsk/+ mice skin (Fig. 4, A and B). Microfibrils from normal mouse skin were well-organized linear arrays (Fig. 4,A). Some microfibrils isolated from Tsk/+ skin were morphologically similar to those of the normal animals, but many other Tsk/+ skin microfibrils were abnormal periodic arrays of beads with indistinct filamentous interbeads and extended periodicity (Fig. 4 B). The axial periodicity of control mice microfibrils and the apparently normal population of microfibrils in Tsk/+ preparations was 55 ± 4 nm, which is comparable to previously published dimensions of microfibrils from other sources (18, 36). However, the abnormal Tsk/+ skin microfibrils had a more extended bead- to-bead distance of 112 ± 11 nm. The number of normal and abnormal microfibrils in triplicate Tsk/+ skin preparations was counted in 200 low-magnification (5,000×) fields; this analysis indicated that the abnormal microfibrils represented 45 ± 10% of the total microfibril population.

Extensive microfibril clusters were also present in the Tsk/+ skin preparations, but not in those from normal skin (Fig. 5, A and B). In these aggregates, intermicrofibril filaments were apparent.

STEM mass analysis in dark field of normal and Tsk/+ skin microfibrils detected a significant mass increase in the Tsk/+ skin microfibril population. Total mass per axial repeat for the normal skin microfibrils was 2,547 kD, whereas that for the abnormal Tsk/+ skin microfibrils was 3,074 kD, an average increase of 527 kD per axial repeat.

Rotary shadowing examination of normal mouse skin microfibrils incubated in the presence of 5 mM EDTA for 10 min demonstrated they had reduced periodicity (38 ± 2 nm), prominent beads, and a diffuse interbead appearance as previously described (35) (Fig. 4,C). The morphologically normal microfibrils from Tsk/+ skin exhibited the same response as the normal mouse skin microfibrils to calcium chelation (Fig. 4,D). However, EDTA-induced changes in the abnormal Tsk/+ skin microfibril population were unequivocally different from those of the apparently normal microfibrils (Fig. 4,D). Axial repeat distances in the abnormal microfibrils were reduced to 61 ± 3 nm, and beads did not appear prominent, although interbeads became more indistinct. EDTA failed to disrupt the large Tsk/+ microfibril aggregates that remained tightly associated but with less well-defined intermicrofibril filaments (Fig. 5, C and D).

To study whether Tsk/+ mice skin microfibrils had altered stability, the effects of chaotropic agents were assessed. 4 M GuHCl-treated normal mice skin microfibrils and the apparently normal Tsk/+ skin microfibrils underwent a conformational change similar to that observed for EDTA, with periodicities reduced to 43.9 ± 9.2 nm, although they remained beaded polymers as previously reported for bovine microfibrils (36). In contrast, the abnormal Tsk/+ microfibrils became disordered networks with loss of overt periodicity (data not shown).

In the presence of calcium, microfibrils isolated from normal and Tsk/+ mice were degraded to short, beaded arrays by trypsin or neutrophil elastase by 18 h. There were no obvious differences in the susceptibility of the control mice microfibrils and the morphologically normal Tsk/+ skin microfibrils to degradation by either enzyme. Short arrays of normal and abnormal Tsk/+ skin microfibrils were detected after digestion, as judged by morphology (data not shown). When EDTA-treated microfibrils were incubated with enzymes, all microfibrils were rapidly degraded within 2 h, with only beaded domains apparent.

A previous study of Tsk mice demonstrated that a tandem duplication within the Fbn 1 is associated with the mouse Tsk mutation and established that the mutant allele is transcribed in Tsk tissues (37). This investigation was undertaken to study whether Tsk fibrillin-1 protein is produced and assembled, and if so, to examine the consequences for microfibril organization.

The Tsk mutation is predicted to give rise to an oversized fibrillin-1 molecule with a calculated molecular mass of 418 kD (37) and an expected presence of seven potential additional N-glycosylation sites. We demonstrate now that cultured dermal fibroblasts of Tsk/+ mice synthesize and secrete normal fibrillin-1 with an apparent molecular mass of 330 kD and a fibrillin-1 molecule with an apparent molecular mass of 450 kD in comparable amounts. The electrophoretic behavior of Tsk fibrillin-1 was in accordance with the predictions and confirmed full expression of the mutant protein. We have several lines of evidence that this oversized molecule aggregates and is incorporated into microfibrils.

Pulse-chase experiments confirmed the incorporation of comparable levels of newly synthesized normal and oversized fibrillin-1 monomers into the extracellular matrix of Tsk dermal fibroblast cultures. Confocal laser scanning microscopy confirmed the presence of a cutaneous microfibrillar network in Tsk mice. Immunoreactivity differences to normal mice using antibodies raised to fibrillin-1 and to intact microfibrils could reflect altered microfibril abundance, changed accessibility of fibrillin-1 epitopes, and/or secondary extracellular matrix changes in affected mice. The transmission electron microscopy study demonstrated that Tsk/+ skin microfibrils were blurred in appearance with no obvious periodicity and limited elastin deposition, indicating structural microfibrillar changes.

Ultrastructural examination of beaded microfibrils isolated from Tsk/+ skin revealed clear evidence for two distinct microfibril populations, one composed of morphologically normal microfibrils, whereas the other population of microfibrils was abnormal with diffuse interbeads and longer than normal periodicity. No other microfibrils with such structural characteristics have previously been demonstrated from any tissue or cellular source. These observations indicate that the abnormal microfibrils compose Tsk fibrillin-1 and have consequently altered molecular organization. The dark-field STEM data demonstrating increased mass of Tsk/+ skin microfibrils support this conclusion. The extra mass presumably reflects incorporation of the larger than normal Tsk fibrillin-1 molecules in the abnormal microfibril pool, although the involvement of other microfibril-associated molecules cannot be excluded (1, 8, 9, 18, 19, 36, 39). The large microfibrillar clusters that further distinguished the Tsk/+ skin microfibril preparations appeared to contain the abnormal microfibrils that presumably have altered surface interactive properties. Structural differences between morphologically normal and abnormal Tsk/+ mice skin microfibrils were confirmed in calcium chelation experiments in which the appearance and periodicity of abnormal Tsk/+ skin microfibrils was fundamentally different from those of the morphologically normal microfibrils in these preparations. Removal of calcium from normal microfibrils and fibrillin-1 molecules causes molecular shortening, increased mass per unit length, and conformational changes (16, 30, 35). Peridiocity of abnormal Tsk/+ microfibrils is reduced, although the beads do not appear prominent like those of EDTA-treated control microfibrils; this distinct response is probably due to misalignment, misfolding, or aberrant cross-linking.

The data obtained with Tsk/+ skin microfibrils show that Tsk fibrillin-1 is produced, assembled, and deposited in the extracellular matrix, and that abnormal beaded Tsk fibrillin-1 microfibrils with longer than normal periodicity and altered morphology and organization are present in Tsk/+ skin. The experiments indicate the occurrence of mutually exclusive normal and mutant fibrillin-1 homopolymeric beaded microfibrils and highlight that microfibril periodicity directly reflects monomer length. The data demonstrate a requirement for fibrillin molecules of similar length for copolymerization into microfibrils. The extended internal sequence of Tsk fibrillin-1 does not preclude homopolymer assembly, and primary nucleation roles for the unchanged amino- and carboxy-terminal sequences in assembly are indicated. Molecular selection based on length suggests that lateral alignment of fibrillin molecules is a crucial assembly step and may precede or facilitate linear polymerization. Current models of fibrillin alignment, based on measured molecular dimensions and antibody epitope mapping (29), or extrapolation of molecular length based on cbEGF-like domain dimensions (6), suggest a parallel alignment of fibrillin monomers, although whether the monomers are unstaggered is not resolved. The extended periodicity that results from polymerization of Tsk fibrillin-1 molecules supports an unstaggered arrangement.

Abnormal Tsk fibrillin-1 microfibrils and aggregated clusters presumably contribute to the Tsk phenotype. Mutations of the human counterpart to the Fbn 1 gene, the FBN 1 locus on chromosome 15q (28), lead to a distinct phenotype, the Marfan syndrome (MFS), an autosomal dominant disorder characterized by pleiotropic connective tissue defects and life-threatening vascular complications. The dominant-negative effects of MFS reflects reduced levels of competent microfibrils in MFS patient tissues and cell cultures (17, 28) and indicates that copolymerization of normal and mutant allele products is universally disruptive to microfibril structure and function. The Tsk fibrillin-1 mutation may not predispose to tissue and vascular alterations precisely because stable heteropolymers of both normal and Tsk fibrillin-1 cannot form, with the normal fibrillin-1 microfibril population able to sustain normal vascular development and functional elastic fiber formation. Altered growth factor binding potential, either through duplicated fibrillin-1 8-cysteine motifs or altered microfibril association with latent transforming growth factor binding proteins (8, 38), may contribute to the characteristic upregulation of extracellular matrix molecules in Tsk mice. It is interesting to speculate that human disorders such as congenital facial dystrophy or restrictive dermopathy (5, 39), characterized by excess collagen deposition and “tight” skin, could have a similar aetiology to the Tsk phenotype.

cbEGF

calcium-binding EGF

cys

cysteine

GuHCL

guanidinium chloride

Fbn 1

fibrillin 1 gene

met

methionine

MFS

Marfan syndrome

STEM

scanning transmission electron microscopy

Tsk

Tight skin mutation