Characterization of mutant type VII collagens underlying the inversa subtype of recessive dystrophic epidermolysis bullosa

Abstract

Background:

Patients with recessive dystrophic epidermolysis bullosa (RDEB) lack functional type VII collagen (C7) leading to skin fragility, bullae, and erosive wounds. RDEB-Inversa (RDEB-I), a subset of RDEB, is characterized by lesions localized to body areas with higher skin temperatures such as flexures and skin folds.

Objective:

We aimed to determine if C7 derived from RDEB-I mutations had structural and functional aberrancies that were temperature sensitive and could be reversed by lowering the temperature.

Methods:

In this study, we generated 12 substitution mutations associated with RDEB-I via site-directed mutagenesis and purified recombinant C7 protein. These C7 mutants were evaluated for structural parameters (trimer formation and protease sensitivity) and the ability to promote keratinocyte migration at 37°C (the temperature of skin folds) and 30°C (the maximum skin temperature of arms and legs). Fibroblasts derived from RDEB-I patients were evaluated for C7 secretion and cellular migration at both temperatures.

Results:

C7s from RDEB-I mutations exhibited decreased thermal stability, increased sensitivity to protease digestion, diminished formation of collagen trimers, and reduced ability to promote keratinocyte migration compared with wild-type C7. In addition, fibroblasts derived from RDEB-I patients demonstrated intracellular accumulation of C7 and abnormal cell migration at 37°C. All of these aberrancies were corrected by reducing the temperature to 30°C. C7s generated from severe-RDEB mutations (non-Inversa) did not display temperature-dependent perturbations.

Conclusion:

These data demonstrate that RDEB-I mutations generate C7 aberrancies that are temperature dependent. This may explain why RDEB-I patients develop clinical lesions in areas where their skin is considerably warmer.

1. Introduction

Recessive dystrophic epidermolysis bullosa (RDEB) is a hereditary disease characterized by skin fragility, blister formation, milia, and scarring [1,2]. It is caused by mutations in the gene (COL7A1) encoding type VII collagen (C7) [3–6]. Differing severities and subtypes of RDEB result from combinations of various mutations, such as premature termination codons (PTCs), missense, and splice-site mutations on both alleles of the COL7A1 gene [1,2,7,8]. One RDEB subtype, RDEB-Inversa (RDEB-I), consists of glycine and arginine substitutions within or bordering the triple helical (TH) domain of C7. These patients develop skin lesions predominantly in areas of the body where the skin is warmer such as the axillae, genital region, mucosae, and skin folds [1,2,9,10], and have been subtyped as the “inverse” form of RDEB.

C7 is a major component of anchoring fibrils (AFs), structures that are required for the adherence of the epidermis to the dermis at the dermal-epidermal junction (DEJ) [11–14]. C7 is a homotrimer composed of three identical alpha chains, each comprised of a central collagenous TH domain consisting of Glycine-X-Y amino acid repeats flanked by a large 145 kD non-collagenous, globular domain (NC1) on the amino-terminus and a smaller 34 kD non-collagenous, globular (NC2) domain on the carboxy-terminus. These alpha chains assemble into homotrimers that form AFs stabilized by disulfide bonding [15–18].

Previously, we have expressed and purified large quantities of recombinant, full length, wild-type C7 and its NC1 domain in human 293 cells and completed extensive structural and functional studies [19–22]. C7 is characterized by (i) its ability to assemble into correctly folded helical trimers that are resistant to protease degradation, (ii) having specific affinity for various extracellular matrix components such as type IV collagen and laminin 332, (iii) the ability to support the matrix attachment of dermal fibroblasts, and (iv) the ability to promote the cellular migration of both fibroblasts and keratinocytes. Using structural and functional assays, we elucidated genotype-phenotype correlations by producing various RDEB substitution mutations. Characterization of individual mutations using these assays has led to a greater correlation between the functional consequences of structural defects and clinical phenotype [21,22].

A previous study identified 19 specific missense mutations associated with RDEB-I and outlined potential genotype-phenotype correlations [9]. C7 homotrimers found in a number of RDEB-I patients have been shown to dissociate at lower temperatures (33–35°C) in vitro compared to normal C7 homotrimers (39–41°C) [23]. Given this information, and the fact that RDEB-I lesions are mainly restricted to body sites with higher skin temperature, it has been hypothesized that the pathophysiology of RDEB-I is temperature dependent [9]. Nevertheless, the structural and functional characteristics of the C7s generated by specific RDEB-I COL7A1 mutations have not been extensively examined. In the current study, we generated 12 specific mutations associated with RDEB-I through site-directed mutagenesis and generated purified mutant RDEB-I type VII collagens (RDEB-I C7s). We compared the structural and functional properties of the purified recombinant RDEB-I C7s with those of purified recombinant wild-type C7 and C7s generated from severe-RDEB mutations (non-Inversa form) at both 37°C and 30°C. We chose these temperatures because 37°C corresponds to the skin temperature in body folds, while 30°C corresponds to the skin temperature on the arms and legs. We also studied the functional characteristics of cultured dermal fibroblast from RDEB-I patients and compared their functional characteristics with cultured fibroblasts from normal individuals. Our results indicate that RDEB-I mutations generate reversible, temperature-dependent, perturbations in the structure and function of C7.

2. Materials and methods

2.1. Cell culture

The human embryonic kidney cell line 293 (ATCC, Rockville, MD) was routinely cultured in Dulbecco’s modified essential medium (DMEM)/Ham’s F12 (1:1) supplemented with 10% fetal bovine serum. Primary human keratinocytes were purchased from Cascade Biologics (Portland, OR) and cultured in low calcium, serum-free keratinocyte growth medium supplemented with bovine pituitary extract and epidermal growth factor (SFM; GIBCO BRL, Gaithersburg, MD) as described by Boyce and Ham and modified by O’Keefe and Chiu [24,25]. Third or fourth passage keratinocytes were used for cell migration studies. Primary RDEB-I and normal human fibroblasts were maintained in DMEM/Ham’s F12 (1:1) supplemented with 10% fetal bovine serum. Cells were passaged as they reached confluence, and all experiments were performed on cells between passages 2–4.

Dermal fibroblasts isolated and banked from skin biopsies of RDEB-I patients as well as from one non-Inversa RDEB patient from our previous study [26,27] were used in this study and placed in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum in 5% CO₂ at 37°C.

2.2. Site-directed mutagenesis and transfection

Site-directed mutagenesis was performed on C7 cDNA in the pRC/CMV vector using a commercial kit (QuikChange™ II site-directed mutagenesis kit, Stratagene Inc., La Jolla, CA) according to the manufacturer’s instructions as described previously [21,22].

The expression vector encoding for wild type or RDEB-I mutant C7 cDNA was used to transfect the human embryonal kidney cell line 293 (ATCC, Rockville, MD) using Lipofectin (Gibco-BRL, Gaithersburg, MD) and stable clones were selected using 500 μg of G418/ml as described [21,22].

2.3. Protein purification and analysis

For immunoblot analysis, clonal cell lines resistant to G418 were grown to confluence, after which the medium was changed to serum-free medium and the cultures were maintained for an additional 24 hrs. The media were collected and subjected to 6% SDS-PAGE followed by immunoblot analysis with polyclonal antibodies to the NC2 domain of C7 [20] and a horseradish peroxidase-conjugated goat anti-rabbit IgG and enhanced chemiluminescence detection reagent (Amersham, UK).

Large-scale purification of recombinant wild type or RDEB-I mutant C7 from serum-free media was accomplished as described previously [21,22].

2.4. Protease digestion

Purified wild type recombinant C7 or RDEB-I mutant C7 was incubated with chymotrypsin (Sigma) at an enzyme-to-substrate ratio of 1:10 by weight in 50 mM Tris-HCl, pH 7.4, 150 mM NaCl at 37°C or 30°C for 4 hours and then analyzed by SDS-PAGE, followed by immunoblot analysis with a rabbit polyclonal antibody recognizing the sequences overlapping the borders of collagenous (P1) and NC2 domains as described [21,22].

2.5. Cell migration assay

Keratinocyte and fibroblast migration assays were performed by the method of Albrecht-Buehler, as modified by Woodley et al. for computerized quantification [28,29]. Briefly, for the keratinocyte migration assays, colloidal gold salts were immobilized on coverslips and coated with purified wild type recombinant C7 (20 μg/ml) or with RDEB-I mutant C7 (20 μg/ml) prior to adding a suspension of keratinocytes to each well. For assessment of patient and normal human fibroblast migration, commercially available type I collagen (30 μg/ml, BD Biosciences) was used to coat the gold salts immobilized on coverslips. In both types of migration assays, the cultured cells were grown at 37°C and then suspended, plated on the coverslips, and allowed to migrate for 16–20 hours at either 30°C or 37°C. The cells were fixed in 0.1% formaldehyde in PBS and examined under dark field optics with a video camera attached to a computer equipped with a frame grabber. The computer analyzes 15 nonoverlapping fields in each experimental condition with NIH Image 1.6 and determines the percentage area of each field consumed by cell migration tracks, a so-called migration index (MI). The experiments were repeated three times.

The methodology to determine statistical differences in MIs between experiments has previously been published [29]. In brief, statistical analyses of the MIs between triplicate sets of experimental conditions were performed using Microsoft Excel. All data were expressed as means ± standard deviation (SD) of separate experiments. Differences between means were determined by the Student t-test for unpaired samples. Only those at p < 0.05 were considered significant and differences at p > 0.05 were insignificant.

3. Results

3.1. Recombinant production of mutant RDEB-I C7s.

In order to examine the underlying molecular defects associated with C7s generated by RDEB-I mutations, we conducted site-directed mutagenesis to generate 12 C7-expression constructs containing mutations associated with RDEB-I within or bordering the TH domain of C7, as shown schematically in Fig. 1A. Transfection of these constructs into 293 cells resulted in secretion of 290-kDa RDEB-I mutant C7s at levels similar to that of wild type C7 (Fig. 1B). From the serum-free culture media of these stably transfected cells, we purified milligram amounts of each recombinant mutant C7 protein following our established protocol [21,22]. As shown in Fig. 1C, all of purified RDEB-I mutant C7s ran as a single band corresponding to the alpha chain of C7, revealed by SDS-PAGE followed by Coomassie Blue staining. These data indicate that the mutant C7s were secreted from cells as full-length 290 kDa intact proteins.

Figure 1. Schematic representation of RDEB-I C7 mutants generated by site directed mutagenesis and expression and purification of recombinant mutant C7 protein.

(A) The full-length (294 amino acids) C7 alpha chain is composed of a large triple-helical (TH) domain flanked on either side by non-collagenous domains, NC1 and NC2. The TH domain contains an exposed 39 amino acid hinge region (delineated in orange) that is sensitive to protease digestion. 12 RDEB-I mutants generated by site directed mutagenesis are shown in the schematic with the approximate positions of the mutations indicated by arrows. (B) Conditioned media from wild type C7 (WT) or 12 RDEB-I mutant C7s as indicated transfected 293 cells were concentrated and subjected to 6% SDS-PAGE followed by immunoblot analysis with an affinity-purified polyclonal antibody to NC1 domain. The positions of the molecular weight marker and 290-kDa C7 are indicated. (C) 6% SDS-PAGE and Coomassie Blue staining of wild type C7 (WT) and RDEB-I mutant C7s purified from conditioned media of stably transfected 293 cells. All samples were run under reducing conditions.

3.2. The structural aberrancies of RDEB-I C7s are temperature dependent.

It has been shown previously that wild-type C7 in vitro will assemble into 900-kDa trimers linked by disulfide bonding [21,22]. Under non-reducing conditions, however, the RDEB-I C7s ran as mixtures of both 290-kDa monomers and 900-kDa trimers. In contrast, wild-type C7 ran purely as a 900-kDa trimer (Fig. 2). These data indicate that RDEB-I mutations generate full-length C7 alpha chains, but that these RDEB-I C7s either have a defective ability to form trimers or there is decreased stability of the trimers once they are formed.

Figure 2. Impaired trimer formation by RDEB-I C7 mutant proteins.

Purified recombinant C7s were subjected to 4–15% SDS-PAGE and immunoblot analysis probed with a polyclonal antibody to the NC2 domain. Proteins were non-reduced. The positions of molecular weight marker, monomers (M) and trimers (T) of C7 are indicated. In the absence of a reducing agent, wild-type C7 is observed exclusively as a ~900 kDa trimer composed of three identical C7 alpha chains (WT). Recombinant RDEB-I C7s as indicated all displayed a diminished ability to form trimers as indicated by a mixture of both monomers (single C7 alpha chain = 290 kDa) and trimers (~900 kDa). This experiment was repeated three times and similar results were obtained in two other independent experiments.

We next evaluated the newly formed RDEB-I trimers for their stability and proper triple helical conformation in the collagenous domain. Purified wild-type and RDEB-I C7s were subjected to partial digestion with chymotrypsin and the digestion products were detected by immunoblot analysis. We also included in these assays two C7s generated from two previously published severe-RDEB mutants, a non-Inversa glycine substitution (G2049E) and a non-Inversa arginine substitution (R2063W). At 37°C, the non-collagenous NC1 and NC2 domains of wild-type C7 were removed. The TH domain of wild-type C7 resisted digestion, except for partial cleavage into P1 and P2 fragments, as previously described [13] and as diagrammed schematically in Fig. 3A. As shown in Fig. 3B, chymotrypsin digestion of wild-type C7 at 37°C produced a 200-kDa TH fragment and a 120-kDa P1 fragment. In contrast, chymotrypsin treatment of all RDEB-I C7s resulted in complete digestion of the TH domain leaving no detectable TH fragment. Chymotrypsin digestion of some RDEB-I C7s (R2063G, R2069C, G2088R, G2689R, G2695S, G2719A, G2775S) produced P1 fragments. In addition, chymotrypsin digestion of the R2063G, R2069C, G2088R, G2689R, and G2695S mutant C7s also revealed one or more additional bands below the P1 fragment. For other RDEB-I C7s (G1907D, G2213R, G2472D, R2622W, R2628W), chymotrypsin digestion yielded neither a TH nor a P1 fragment. Protease digestion of the two non-Inversa severe-RDEB mutant C7s (G2049E and R2063W) generated only a P1 fragment at 37°C.

Figure 3. Ability of RDEB-I C7s to resist protease digestion is temperature dependent.

(A) A schematic of the C7 alpha chain. Proteases (labeled P in the schematic), cleave the NC1 and NC2 domains from the triple helical domain (TH). The TH domain of C7 contains a 39 amino acid non-helical interruption (hinge region) where proteases cleave the TH domain almost in half and generate the carboxy-terminal P1 and the amino-terminal P2 fragments. Arrows indicate the protease cleavage sites. As indicated, the α-TH antibody recognizes the P1 fragment or full TH domain but not the P2 fragment. (B and C) Purified wild-type C7 (WT), severe-RDEB C7s (indicated by^✦) or RDEB-I C7s, as indicated, were treated with chymotrypsin at 37°C (B) or 30°C (C) and analyzed by 6% SDS-PAGE followed by immunoblot labeling with a polyclonal antibody to the α-TH/NC2 small domain overlapping the borders of TH and NC2 domains. The positions of molecular weight markers, the 200-kDa intact triple helical domain (TH) and the 120-kDa carboxyl-terminal half of the TH fragment (P1) are indicated. This experiment was repeated three times and similar results were obtained in two other independent experiments.

Next, we assessed whether the stability of these trimers was affected by lowering the temperature from 37°C to 30°C during the protease digestion assays. We carried out protease digestions at 30°C, the reported maximum skin temperature (27°C - 30°C) of hands, forearms and lower legs [30,31]. Interestingly, most of the RDEB-I C7s (R2063G, R2069C, G2472D, R2622W, R2628W, G2689R, G2695S, G2719A, G2775S) showed a new digestion pattern nearly identical to that of wild-type C7 with two bands representing the TH and intact P1 fragments as detected by immunoblot analysis (Fig. 3C). Two RDEB-I C7s (G1907D, G2088R) still exhibited complete digestion of the TH fragment and only the P1 fragments were detected at 30°C. In contrast, the digestion pattern of the two non-Inversa, severe-RDEB mutants revealed only P1 fragments at 30°C, unchanged from that seen at 37°C. These results demonstrate that the TH domain of all RDEB-I C7s are less stable against proteolysis than that of wild-type C7 at 37°C. Furthermore, this instability was reversed at 30°C, the cooler temperature of skin found on the arms and legs.

3.3. The functional aberrancies of RDEB-I C7s are temperature dependent.

We have shown previously that wild-type C7 strongly promotes human keratinocyte migration [22]. In order to test whether perturbations in the folding of RDEB-I C7s have functional consequences, we examined the migration of human keratinocytes juxtaposed to recombinant RDEB-I C7s, severe-RDEB C7s (non-inverse), or wild-type C7 at 37°C and 30°C (representing skin temperature in skin folds versus skin temperature one extremities, respectively) using a colloidal gold salt migration assay developed by Albrecht-Buehler and modified for computerization [28,29]. Fig. 4A shows representative tracks from a number of RDEB-I C7s, severe-RDEB C7s (non-inverse), and wild-type C7 at 37°C and 30°C. Fig. 4B shows the measurement of the MIs of the keratinocytes juxtaposed to the different types of C7s. Wild-type C7 promoted human keratinocyte migration similarly at both temperatures. In contrast, 11 of 12 RDEB-I C7s (with the exception of G2213R) exhibited decreased ability to promote keratinocyte migration at 37°C. When placed at 30°C, all RDEB-I C7s (except G2088R) demonstrated enhanced activity in promoting keratinocyte migration to levels 80–100% of that of wild-type C7 (Fig. 4B). In contrast, two severe-RDEB mutants (G2049E and R2063W) failed to promote migration above 20% of that seen in wild-type C7 at both temperatures. These data demonstrate that the potent pro-motility function of C7 on human keratinocytes is indeed impaired as a function of increasing temperature in a majority of RDEB-I mutant C7s. Table I summarizes the structural and functional aberrancies of RDEB-I C7s when compared with wild-type C7.

Figure 4. Temperature-dependent migration of human keratinocytes on RDEB-I C7s.

(A) Coverslips were coated with colloidal gold salts, then keratinocytes were plated on either RDEB-I C7s, severe-RDEB (indicated by^✦) or wild-type C7 (20 μg/ml) and incubated at either 37°C or 30°C for 18 hours. Representative fields were photographed at 40X under dark field optics. (B) The Migration Index (MI) expresses the percentage of the total microscopic field area consumed by the areas of the black migration tracks. Each bar represents human keratinocytes migrating on a purified RDEB-I C7s or severe-RDEB^✦ at either 37°C (red bars) or 30°C (blue bars) and represents the percentage of cell migration compared with wild-type C7. ^‡ For wild-type C7 (WT), migration at 30°C represents the percentage of cell migration compared to with wild-type C7 at 37°C (set at 100%). Error bars represent the mean +/− SD of MI, 15 non-overlapping fields with at least 2–4 cells in each field were counted from each replicate experiment and repeated three times, *P:<0.05, **P:<0.01 between MI at 37°C and 30°C.

Table I:

Summary of Wild-Type Versus RDEB-I C7 Structure and Function

RDEB-I Mutants	Trimer Formation	Protease Stabilitya (30°C /37°C)	Cell Migrationb (30°C/37°C)
Wild Type	T	Resistant/Resistant	+++/+++
G1907D	T+M	Sensitive/High Sensitive	+++/+
R2063G	T+M	Resistant/Sensitive	+++/+
R2069C	T+M	Resistant/Sensitive	+++/++
G2088R	T+M	Sensitive/Sensitive	+++/+
G2213R	T+M	Sensitive/High Sensitive	+++/++
G2472D	T+M	Resistant/High Sensitive	+++/++
R2622W	T+M	Resistant/High Sensitive	+++/++
R2628W	T+M	Resistant/High Sensitive	+++/+
G2689R	T+M	Resistant/Sensitive	+++/++
G2695S	T+M	Resistant/Sensitive	+++/++
G2719A	T+M	Resistant/Sensitive	+++/++
G2775S	T+M	Resistant/Sensitive	+++/+

^aResistant, Digestion pattern with TH and P1 fragments as seen in wild type; Sensitive, digestion with only P1 fragment; High Sensitive, neither TH or P1 fragments detected

^bPercentage of MI to wild type: +++, 80%–100%; ++, 50–80%; +< 50%

3.4. Temperature-dependent cellular aberrancies in RDEB-I fibroblasts.

To determine whether RDEB-I mutations also cause aberrancies in RDEB-I skin cells, we characterized fibroblasts from two RDEB-I patients from our previous studies under different temperatures [26]. Patient 1 was heterozygous for G1907D and a PTC mutation, and Patient 2 was heterozygous for R2069C and a frameshift mutation resulting in a PTC. Normal human fibroblasts (NHF) and RDEB-I fibroblasts were incubated in serum-free medium at either 37°C or 30°C for 48 hours. As shown in Fig. 5A, at 37°C and 30°C, RDEB-I fibroblasts and NHFs had similar levels of intracellular C7, when subjected to immunoblot analysis. Likewise, NHFs were able to secrete C7 extracellularly into the conditioned medium at both 37°C and 30°C. In contrast, there was no detectable amount of secreted RDEB-I mutant C7s (G1907D and R2069C) into the medium at 37°C. Nevertheless, this impairment of C7 secretion by RDEB-I fibroblasts was reversed when the incubation temperature was lowered to 30°C. When we performed densitometry analysis to evaluate the C7 expression normalized to β-tubulin, amounts of C7 determined from cellular extract in PT1 and PT2 were 56.5 and 90% of that seen in normal human fibroblasts at 37°C. Because there were not increased amounts of mutant C7 accumulated intracellularly within the RDEB-I fibroblasts, the lack of C7 detected in the condition medium was likely due to rapid degradation of the secreted mutant C7s that were generated at 37°C. To exclude the possibility that temperature-dependent changes in secreted C7s results from reduced C7 production, we have performed an experiment using fibroblasts from one non-Inversa RDEB mutant and showed that there was reduced amounts of C7 mutant detected in cellular extract compared with wild type C7 seen in normal human fibroblasts. However, the amounts of C7 mutant detected in medium is similar to the wild type C7 seen in normal human fibroblasts at both temperatures (Supplementary Fig. 1). These data indicate that temperature-dependent changes in secreted C7s are a unique feature of RDEB-I mutants and are not due to an effect of the amounts of C7 production itself.

Figure 5. Characterization of primary RDEB-I fibroblasts.

(A) Total cellular lysates and conditioned media were prepared from cultured fibroblasts from two RDEB-I patients and from normal human subject (NHF) incubated at confluence for 48 hrs at either 37°C or 30°C. Samples were separated by 4–15% SDS-PAGE and blotted with a polyclonal NC1 antibody (β-tubulin is used as a loading control for lysates). (B) NHF and fibroblasts from two RDEB-I patients (Pt1 and Pt2) were subjected to a migration assay using collagen I as the substratum at either 37°C or 30°C. Panels are representative fields photographed at 40X under dark field optics showing black cell migration tracks. (C) Bar graphs show computer-generated Migration Indices (MIs) at 30°C (blue bars) and at 37°C (red bars). This experiment was repeated three times and similar results were obtained in two other independent experiments. Error bars represent mean +/− SD. *P:<0.05 between MI at 37°C and 30°C.

As previously reported, RDEB fibroblasts have abnormal cellular hypermotility likely due to poor substratum adherence secondary to a paucity of functional C7 [32]. To determine whether the RDEB-I cells exhibited a similar abnormal hypermotile cellular phenotype, we subjected RDEB-I fibroblasts and normal fibroblasts to the colloidal gold salt migration assay. In this fibroblast migration assay, type I collagen is coated and immobilized on the colloidal gold particles. Type I collagen is a matrix known to promote fibroblast motility. As shown in Fig. 5B and 5C, RDEB-I fibroblasts were hypermotile when juxtaposed to the type I collagen at 37°C, as evidenced by larger, irregular amorphous migration tracks compared with those of NHF control cells. When the temperature of the assays was lowered to 30°C, however, the RDEB-I fibroblasts reduced their motility to levels similar to that of NHFs.

4. Discussion

RDEB-I has been classified as a distinct subtype of RDEB for four decades, yet much remains unknown about the biochemical malfunctions or molecular mechanisms that lead to the distinct clinical features of RDEB-I and its tendency to have lesions in areas of skin folds where the skin is warmer than other anatomic skin sites. In this study, we sought to determine whether the structural and functional properties of C7s or cells derived from known RDEB-I genetic mutations are temperature-dependent. Using site-directed mutagenesis, we engineered 12 stable 293 cell lines overexpressing mutant C7s corresponding to published mutations associated with RDEB-I patients. Recombinant C7s were purified from these cells and compared to wild-type C7 and severe-RDEB C7s (non-inverse) in a number of structural and functional assays that tested the ability of the C7s to form trimers, resist protease digestion, and promote keratinocyte migration. Our results demonstrate that RDEB-I C7s are indeed thermolabile. At 37°C, these RDEB-I C7s become increasingly sensitive to protease digestion, display impaired abilities to assemble into trimers, and show reduced ability to promote keratinocyte migration. At the lower temperature of 30°C, these abnormalities were largely reversed. Abnormal RDEB-I fibroblast C7 secretion and cell motility at 37°C was additionally shown to be correctable at the lower temperature of 30°C.

The specificity of the temperature-dependent structural and functional properties of RDEB-I C7 mutants used in this study was further supported by analysis of two severe-RDEB (non-Inversa) C7 mutants, G2049E and R2063W that we generated from our previous study [22]. Both of these severe-RDEB C7 mutants were sensitive to protease digestion and failed to promote cell migration at both temperatures. It is interesting to note that the arginine at codon 2063 was found substituted by glycine in RDEB-I (R2063G), whereas it was substituted by tryptophan (R2063W) in severe-RDEB. Specifically, this substitution was in exon 73, in close vicinity to a non-collagenous 39 amino acid helical interruption (the so-called “hinge region”). Arginine is an aliphatic, hydrophilic, and positively charged amino acid present at the third position of a Glycine-X-Y motif in the TH of C7. It is therefore conceivable that replacement of arginine with tryptophan (R2063W), seen in non-Inversa, severe-RDEB, significantly alters the conformation of the C7 molecule and induces more unraveling in the area near the hinge region, increasing accessibility and vulnerability to protease degradation. In the case of R2063G associated with RDEB-I, the substitution of arginine with glycine may cause less conformational changes than a substitution with tryptophan (R2063W), as seen in severe-RDEB. Thus, milder change in R2063G may facilitate the restoration of proper conformation and function in response to lower temperature.

The biological consequences of glycine substitutions in C7 can be numerous and consequential to the formation and function of AFs via different mechanisms. In general, the tight packing of collagen triple helices provides relative resistance to protease degradation. Glycine substitutions within the collagen tripeptide repeat can alter this tightly packed conformation and subsequently alter the C7’s function and/or stability. A local disruption in the triple helical structure can be detected by measuring the susceptibility of the collagen to proteolytic cleavage in vitro. In the current study, 8 of 12 RDEB-I C7 mutants we generated involved glycine substitutions. Many of these RDEB-I C7s demonstrate temperature-dependent thermostability to protease digestion. All mutants, except G2088R, exhibited enhanced sensitivity to protease digestion at 37°C but regained resistance or partial resistant at 30°C. Two RDEB-I C7 mutants (G1907D and G2213R) were still partially sensitive toward protease digestion at 30°C, indicating that these mutant C7s may constitute less stable RDEB-I C7s and may require even lower temperatures to be completely protected from protease digestion. Interestingly, the four RDEB-I C7s associated with arginine substitutions studied here (R2063G, R2069C, R2622W, and R2628W) were all prone to protease digestion at 37°C but regained full resistance at 30°C, mirroring the wild-type C7 digestion pattern at the lower temperature. It is possible that compared with glycine substitutions, the conformational molecular C7 changes induced by arginine substitutions are less.

We believe that the temperatures selected for the analyses herein are physiologically relevant. Utilizing digital thermography, one study previously demonstrated higher body temperatures in the same sites typically affected by RDEB-I: proximal flexures, neck, lower back, and sub-mammary folds [9]. Another study of bathing suit ichthyosis used a similar experimental approach to show a correlation between scaling distribution and warmer body sites (>33–34°C) such as the antecubital fossae, popliteal fossae, axillae, and neck [33]. The typical core body temperature, as well as that of the oral cavity and axilla, is approximately 37°C. Therefore, this temperature is a good upper threshold for testing migration and structure of RDEB-I cells and function of RDEB-I C7s. The temperature of 30°C was assessed because this is the average skin temperature of the hands, forearms, and lower legs – skin sites less commonly affected by bullae and erosions in RDEB-I patients. In our experiments, we found that the skin cells from normal individuals migrated identically at 30°C and 37°C. The fact that the majority of recombinant RDEB-I C7s were shown to be functional at 30°C in promoting keratinocyte migration and correspondingly dysfunctional at 37°C lends strong support to the temperatures selected. However, one mutant (G2213R) maintained migration function at 37°C suggesting that this mutant C7 is likely the most stable of the RDEB-I C7s examined in this study.

The implications for therapy based on the results presented in this paper are straight forward. Once a confirmed diagnosis of RDEB-I is made, measures to keep the patient in a cool environment may be helpful. Strategies such as using cooled breast or bottled milk, lighter bandaging, less restrictive clothing, and air conditioning might decrease the induction of new blisters. Moving the family to an area that has cooler temperatures might also be beneficial. These common-sense strategies, however, have not been adequately confirmed by evidence-based medicine due to the rarity of the disease and lack of an appropriate comparator as a control.

RDEB-Inversa Highlights.

• Patients with RDEB lack functional Type VII collagen (C7)

• The Inversa subtype (RDEB-I) affects body areas of higher temperature

• The structural and functional abnormalities of RDEB-I C7 are temperature-dependent

• The abnormal RDEB-I fibroblast phenotype is correctable at lower temperature

Abbreviations:

NC1 and NC2

N- and C- terminal non-collagenous domain, respectively, of type VII collagen

C7

type VII collagen

AF

anchoring fibril

PTC

premature termination codon

DEJ

dermal-epidermal junction

RDEB

recessive dystrophic epidermolysis bullosa

RDEB-I

Inversa type of RDEB

TH

triple helical

NHF

normal human fibroblasts

kDa

kilodalton

PAGE

polyacrylamide gel electrophoresis

MI

migration index