兽研[内容请查看原文]
引言
健康的皮肤构成了机体与环境之间的外部边界，对机械应力具有很强的抵抗力。这种抵抗力的关键结构包括基底角质形成细胞中的中间丝、真皮-表皮交界处的半桥粒以及基底膜区的锚定原纤维（Uitto and Richard 2005）。这些结构中任何一个的缺陷都可能导致大疱性表皮松解症（EB），这是一组遗传性水疱性疾病，其特征是轻微创伤后出现真皮-表皮组织分离。根据皮肤裂隙的层次，EB被分为四种主要类型：单纯型EB、交界型EB（JEB）、营养不良型EB和Kindler型EB。迄今为止，人类中至少有16个基因的致病变异与不同形式的EB相关（Bardhan et al. 2020; Has et al. 2020）。
JEB是由介导基底角质形成细胞与下方基底膜之间黏附的蛋白质缺陷所引起的。层粘连蛋白-332由层粘连蛋白α3、β3和γ2链组成，是基底膜中将表皮锚定于真皮的关键结构成分。参与该黏附复合物的其他蛋白质包括XVII型胶原蛋白和α6β4整合素。影响编码这些蛋白质基因的致病变异会导致真皮-表皮交界处分离，这是JEB的特征性组织病理学特征（Bardhan et al. 2020; McMillan et al. 2003）。
自然发生的JEB已在多种家养动物物种中有报道，包括犬、猫、牛、马和羊。分子研究已在与人类疾病相关的多个相同基因中鉴定出致病变异，即LAMA3（Capt et al. 2005; Graves et al. 2009; Herrmann et al. 2021; Sartelet et al. 2015）、LAMB3（Kiener et al. 2020; Letko et al. 2025）、LAMC2（Mömke et al. 2011; Murgiano et al. 2015; Spirito et al. 2002）、COL17A1（Fussell et al. 2026; Kiener et al. 2023）、ITGA6（Boussaha et al. 2023）和ITGB4（Fabre et al. 2021; Peters et al. 2015; Suárez‐Vega et al. 2015）。这些发现表明，JEB的遗传基础在人类和家养动物物种之间高度保守（Medeiros and Riet‐Correa 2015）。
在本研究中，我们描述了三只随机繁殖幼犬中JEB的临床、组织病理学和分子特征。利用全基因组测序和靶向分子分析，我们鉴定出LAMA3中一个纯合外显子SINE插入是该疾病可能的遗传原因。
[内容请查看原文]
材料与方法
[内容请查看原文]
动物
本研究纳入了三只患有JEB的随机繁殖幼犬。此外，其中两只幼犬未受影响的母亲和一只临床未受影响的同窝仔犬可用于遗传分析。使用了一组1538个公开可得的、来自遗传多样性犬的全基因组序列进行变异过滤（表S1）。纳入了一只无关对照犬用于靶向基因分型。
[内容请查看原文]
临床和组织病理学检查
所有患病犬均接受了临床检查。从多个解剖部位的病变皮肤采集皮肤活检样本，并按常规处理进行组织病理学检查。进行组织病理学评估以确认大疱性表皮松解症的诊断并表征微观病变。
[内容请查看原文]
全基因组测序和变异检测
研究中所有犬的基因组DNA均使用Maxwell RSC全血DNA试剂盒在Maxwell RSC仪器（Promega，Dübendorf，瑞士）上从EDTA血中提取。对其中一只患病雌犬进行了如下全基因组测序。使用Illumina TruSeq DNA PCR-Free Library Prep Kit（Illumina，San Diego，CA，USA）按照制造商方案从基因组DNA制备无PCR文库。文库在瑞士伯尔尼大学下一代测序平台上使用Illumina NovaSeq 6000仪器（Illumina，San Diego，CA，USA）以双端2 × 150 bp模式（在S4流动槽上与其他文库混合）进行测序。测序产生了2.583亿个读对。变异检测遵循GATK最佳实践工作流程，在我们内部的Nextflow DSL2流程DogWGS（https://github.com/vetsuisse‐unibe/DogWGS）中实现。每个样本的读段使用fastp [v0.23.4]进行接头和质量修剪，使用BWA‐MEM2 2.2.1比对到德国牧羊犬基因组组装（UU_Cfam_GSD_1.0，GCF_011100685.1；Wang et al. 2021）的修改版本，并补充了来自拉布拉多寻回犬（ROS_Cfam_1.0，GCF_014441545.1）的三条Y染色体序列。比对结果按坐标排序（SAMtools v1.13），PCR/光学重复使用Picard MarkDuplicates（v2.21.8）标记。在碱基质量分数重校准（BQSR）后，使用GATK v4.2.6.1 HaplotypeCaller（GVCF模式）对每个样本进行变异检测，跨样本联合基因分型（GenomicsDBImport → GenotypeGVCFs），过滤（VQSR/硬过滤）并使用snpEff（v5.0e）注释。99.91%的测序读段比对到参考序列，98.78%正确配对，重复率为5.70%，平均错配率为0.63%。去重后的比对在≥ 1×覆盖度下覆盖了参考序列的98.27%，平均覆盖度为29.3×。该基因组已存入欧洲核苷酸档案库，登录号为SAMEA10833662。
[内容请查看原文]
变异过滤
从包含患病犬和1538个公开可获得的犬全基因组序列的多样本、经snpEff注释的队列VCF（表S1）开始，将snpEff注释展平为每行一个转录本效应，生成一个制表符分隔的表格，包含变异坐标、等位基因、每样本基因型和预测的功能后果。由于遗传模式未知，同时考虑了显性和隐性疾病模型。使用自定义流程提取患病犬特有的候选变异。仅保留病例为杂合（0/1）或纯合变异（1/1）且所有其余动物为纯合参考（0/0）或缺失（./.）的变异。随后对特有变异进行过滤，仅保留蛋白质改变变异，即SnpEff预测影响为HIGH或MODERATE的变异。最后，根据剩余变异在16个已知大疱性表皮松解症候选基因中的存在情况进行优先级排序（Has et al. 2020）。
[内容请查看原文]
变异验证和靶向基因分型
候选变异LAMA3:XM_038543644.1:c.5690_5691ins240通过Sanger测序验证，并通过片段长度分析对所有患病犬、可获得的家庭成员和一只无关对照犬进行基因分型。使用QIAGEN Multiplex PCR Kit（Qiagen, Hilden, Germany）按照制造商说明进行PCR扩增，退火温度为60°C，35个扩增循环。正向引物为5′-TCA TGG ATT AAA AAT GGA TGG TC-3′，反向引物为5′-TGC ATG GTT TCA AAT TCA CG-3′。PCR产物经外切核酸酶I和碱性磷酸酶处理后，在5200 Fragment Analyzer（Agilent, Basel, Switzerland）上进行片段长度分析。纯化的PCR产物额外在ABI 3730 DNA Analyzer（Thermo Fisher Scientific, Waltham, MA, USA）上测序，以确定插入SINE元件的精确序列。使用Sequencher 5.1软件（GeneCodes, Ann Arbor, MI, USA）分析序列色谱图。
[内容请查看原文]
结果
[内容请查看原文]
临床和组织病理学检查
研究了两只雌性和一只雄性随机繁殖幼犬。其中两只患病犬为同窝仔，而与第三只幼犬的确切关系未知。所有三只患病幼犬具有相似的临床症状，约在6-8周龄时开始，包括指甲完全脱落、爪垫和口腔黏膜深部溃疡、耳廓内侧水疱和溃疡、腹股沟区和腹部溃疡和大疱，以及釉质发育不全（图1a-c）。其中两只幼犬还患有不愈合的单侧非常大的角膜溃疡（图1d）。所有三只患病幼犬在诊断后几周内因严重临床恶化、食欲丧失和无法控制的疼痛而被安乐死。
[图：临床表型。(a) 患病幼犬腹部多灶性溃疡。(b) 严重的深部爪垫溃疡。(]
EB的诊断通过病变皮肤活检确认。最一致的发现是表皮-真皮分离。伴随一系列反应性变化，取决于病变持续时间和继发性细菌感染的存在。大多数样本中相邻区域可见具有不同显微镜特征的病变。在急性病变中，表皮高度正常或中度增生，表皮-真皮分离形成的间隙中常有纤维蛋白和/或血液积聚（图2a）。真皮变化包括血管扩张、成纤维细胞和血管细胞肥大以及轻度中性粒细胞浸润。在亚急性病变中，中度增生的表皮与真皮分离，真皮有轻度浅表纤维化（图2b）。在伴有继发性细菌感染和/或复发证据的长期病变中，表皮增生、真皮纤维化和炎症更为明显，并存在溃疡、浆细胞性结痂和表皮再生（图2c）。偶尔，中性粒细胞聚集在抬起的表皮下方（图2d）。
[图：患病幼犬皮肤活检的组织病理学发现。(a) 急性病变。表皮与]
[内容请查看原文]
遗传分析
将患病犬的WGS数据与1538个公开可获得的犬基因组进行比较，鉴定出214个杂合和60个纯合特有蛋白质改变变异（表S2）。对16个已知EB候选基因的筛查（Has et al. 2020）鉴定出LAMA3中的单个候选变异。虽然最初被判定为杂合，但短读长比对的视觉检查确定该变异实际上是LAMA3基因中的纯合外显子SINE插入（XM_038543644.1:c.5690_5691ins240；图3a,b）。
[图：LAMA3变异的详细信息。(a) LAMA3基因基因组组织结构示意图，显示内含子]
Sanger测序证实LAMA3第45外显子存在一个240 bp的SINE插入，并确定了其精确序列和插入位点（文件S1）。该插入两侧为16 bp的靶位点重复，并含有一段poly(A)尾。这两个特征从短读长比对中也可明显看出（图3b和文件S1）。序列分析表明该插入为框内插入，预测会在层粘连蛋白α3的卷曲螺旋结构域内引入额外80个氨基酸，而不引入提前终止密码子（XP_038399572.1:p.(Lys1897_Glu1898insXaa[80])）。
片段长度分析表明，在所研究的家系中基因型与表型完全一致。三只患病幼犬均为该插入的纯合子，而其未患病的母犬和未患病的同窝仔犬为杂合携带者。无亲缘关系的对照犬为参考等位基因的纯合子（图3c,d）。
[内容请查看原文]
讨论
在本研究中，我们在LAMA3中鉴定出一个新的纯合外显子SINE插入，可能是三只随机繁殖幼犬发生交界型大疱性表皮松解症（JEB）的原因。该变异在1538份公开可获得的犬全基因组序列中不存在，并且是在一只JEB患病犬的WGS数据中鉴定到的、已知EB候选基因中唯一改变蛋白质的私有变异。该变异的基因型在所研究家系内与疾病表型共分离。应用人类诊断中解释序列变异的ACMG/AMP标准（Richards等，2015）或AVCG v2标准（Desmet等，2026），我们将该变异分类为致病性（表S3）。
LAMA3在人类和动物中均是研究较为充分的JEB候选基因（Capt等，2005；Graves等，2009；Has等，2020；Herrmann等，2021；Kivirikko等，1995；Sartelet等，2015）。它编码层粘连蛋白α3链，与分别由LAMB3和LAMC2编码的β3链和γ2链共同组成层粘连蛋白-332的三个亚基之一。层粘连蛋白-332是基底膜的关键组分，将上皮细胞锚定于下方组织。其组装始于内质网，β3链和γ2链先形成异二聚体，随后α3链的并入完成成熟的异三聚体。因此，α3链对于层粘连蛋白-332的有效组装和分泌至关重要，影响三个亚基中任何一个的致病性变异都会破坏该复合物的完整性并损害真皮-表皮黏附（Bardhan等，2020；Parzymies，2026；Sugawara等，2008）。
假设该SINE插入不改变pre-mRNA剪接（本研究未对此进行研究），预测该插入可维持开放阅读框，并在层粘连蛋白α3的卷曲螺旋结构域中引入额外80个氨基酸。预测的插入包含一段由SINE插入的poly(A)尾编码的15个苯丙氨酸。这样的氨基酸序列显然与卷曲螺旋结构域的七肽重复组织不相容（Lupas，1996）。卷曲螺旋结构域介导与β3链和γ2链形成异三聚体，因此对正常的层粘连蛋白-332组装至关重要（Domogatskaya等，2012）。尽管该插入的功能后果仍有待确定，但这一高度保守结构域内的大幅框内扩展预计会干扰正常的卷曲螺旋形成，从而严重损害层粘连蛋白-332的组装和/或分泌。
所鉴定变异预测的功能后果与患病幼犬中观察到的严重临床表型一致。在人类中，严重的泛发性JEB通常与LAMA3、LAMB3或LAMC2的双等位基因功能缺失变异相关，而允许残留层粘连蛋白-332表达的变异通常导致较轻的表型（Parzymies，2026）。同样，Lama3−/−裸鼠在出生后不久即出现广泛的皮肤水疱，表现出异常的半桥粒、表皮锚定功能丧失，并在出生后数日内死亡（Abrass等，2006；Yao，2017）。
一个有趣的技术发现是，该致病变异最初被GATK HaplotypeCaller报告为杂合53 bp插入，而非纯合240 bp SINE插入。对短读长比对的目视检查发现，跨越断点处有大量软剪切读长和不一致读对，提示存在大得多的插入。由于SINE元件在犬基因组中以多拷贝存在，来源于插入元件内部的读长无法唯一比对，因此获得较低的比对质量评分，而跨越独特侧翼序列的读长则以高置信度比对。变异识别软件对低质量读长的过滤造成了人为的等位基因失衡，导致错误的杂合基因型判定。这一观察说明了标准短读长变异识别流程对重复元件插入的局限性，并强调了目视检查候选变异的重要性，尤其是在评估可能的致病性结构变异时。
尽管随机繁殖的犬常被认为具有遗传多样性，但由于近亲交配（如兄弟姐妹之间或亲子之间）可能导致高水平的近交（Leeb等，2023）。本研究中观察到的分离模式与常染色体隐性遗传一致，表明遗传性疾病也发生于此类群体中，强调了在非纯种动物中考虑遗传病的重要性。该研究进一步扩展了导致犬交界性大疱性表皮松解症的LAMA3致病性变异谱，并证明了全基因组测序结合结构变异的仔细解读在兽医患者遗传性皮肤病诊断中的价值。
作者贡献
Sarah Kiener：研究实施、可视化、写作——原始草稿、写作——审阅与编辑。Ronnie Kaufmann：概念构思、研究实施、可视化、写作——原始草稿、写作——审阅与编辑。Tosso Leeb：概念构思、可视化、写作——原始草稿、写作——审阅与编辑、资金获取。Ori Brenner：概念构思、研究实施、可视化、写作——原始草稿、写作——审阅与编辑。Vidhya Jagannathan：数据管理、写作——审阅与编辑。
资助
本研究由瑞士国家科学基金会资助（资助号310030_200354）。
伦理声明
所有检查和动物实验均在获得所有者书面知情同意后，依照当地法律、法规和伦理指南进行。犬血液样本的采集获得州动物实验委员会批准（瑞士伯尔尼州；许可号BE 71/19和BE94/2022）。皮肤钻孔活检出于诊断目的进行，获得所有者知情同意，符合伦理指南和经批准的程序。
利益冲突
作者声明无利益冲突。
支持信息
[内容请查看原文]
[内容请查看原文]
[内容请查看原文]
[内容请查看原文]
查看英文原文▾
[内容请查看原文]
Introduction
Healthy skin forms the external boundary between the organism and the environment and is very resistant to mechanical stress. Key structures for this resilience are intermediate filaments in the basal keratinocytes, the hemidesmosomes at the dermal‐epidermal junction, and the anchoring fibrils in the basement membrane zone (Uitto and Richard 2005). Defects in any of these structures can result in epidermolysis bullosa (EB), a group of inherited blistering disorders characterized by dermo‐epidermal tissue separation following minimal trauma. Based on the level of skin cleavage, EB is classified into four major types: EB simplex, junctional EB (JEB), dystrophic EB, and Kindler EB. To date, pathogenic variants in at least 16 genes have been associated with the different forms of EB in humans (Bardhan et al. 2020; Has et al. 2020).
JEB results from defects in proteins that mediate adhesion between basal keratinocytes and the underlying basement membrane. Laminin‐332, composed of the laminin α3, β3, and γ2 chains, is a key structural component of the basement membrane that anchors the epidermis to the dermis. Additional proteins involved in this adhesion complex include collagen XVII and the α6β4 integrin. Pathogenic variants affecting genes coding for these proteins result in separation at the dermal‐epidermal junction, the characteristic histopathological feature of JEB (Bardhan et al. 2020; McMillan et al. 2003).
Naturally occurring JEB has been reported in several domestic animal species, including dogs, cats, cattle, horses, and sheep. Molecular studies have identified pathogenic variants in several of the same genes implicated in human disease, namely LAMA3 (Capt et al. 2005; Graves et al. 2009; Herrmann et al. 2021; Sartelet et al. 2015), LAMB3 (Kiener et al. 2020; Letko et al. 2025), LAMC2 (Mömke et al. 2011; Murgiano et al. 2015; Spirito et al. 2002), COL17A1 (Fussell et al. 2026; Kiener et al. 2023), ITGA6 (Boussaha et al. 2023), and ITGB4 (Fabre et al. 2021; Peters et al. 2015; Suárez‐Vega et al. 2015). These findings demonstrate that the genetic basis of JEB is highly conserved across humans and domestic animal species (Medeiros and Riet‐Correa 2015).
In this study, we describe the clinical, histopathological, and molecular characterization of JEB in three random‐bred puppies. Using whole‐genome sequencing and targeted molecular analyses, we identified a homozygous exonic SINE insertion in LAMA3 as the likely genetic cause of the disease.
[内容请查看原文]
Materials and Methods
[内容请查看原文]
Animals
Three random‐bred puppies with JEB were included in this study. In addition, the unaffected mother of two of the puppies and one clinically unaffected littermate were available for genetic analyses. A cohort of 1538 publicly available whole‐genome sequences from genetically diverse dogs was used for variant filtering (Table S1). One unrelated control dog was included for targeted genotyping.
[内容请查看原文]
Clinical and Histopathological Examination
All affected dogs underwent clinical examination. Skin biopsies were collected from lesional skin at several anatomical sites and routinely processed for histopathological examination. Histopathological evaluation was performed to confirm the diagnosis of epidermolysis bullosa and characterize the microscopic lesions.
[内容请查看原文]
Whole Genome Sequencing and Variant Calling
Genomic DNA from all dogs included in the study was isolated from EDTA blood with the Maxwell RSC Whole Blood DNA Kit using a Maxwell RSC instrument (Promega, Dübendorf, Switzerland). One of the affected female dogs was whole‐genome sequenced as follows. A PCR‐free library was prepared from genomic DNA using the Illumina TruSeq DNA PCR‐Free Library Prep Kit (Illumina, San Diego, CA, USA) following the manufacturer's protocol. The library was sequenced on an Illumina NovaSeq 6000 instrument (Illumina, San Diego, CA, USA) in paired‐end 2 × 150 bp mode (within a pool of other libraries on an S4 flow cell) at the Next Generation Sequencing Platform, University of Bern, Switzerland. The sequencing produced 258.3 million read pairs. Variant calling followed the GATK best‐practices workflow as implemented in our in‐house Nextflow DSL2 pipeline DogWGS (https://github.com/vetsuisse‐unibe/DogWGS). Per‐sample reads were adapter and quality‐trimmed with fastp [v0.23.4], aligned to a modified version of the German Shepherd Dog genome assembly (UU_Cfam_GSD_1.0, GCF_011100685.1; Wang et al. 2021), supplemented by three Y chromosome sequences from a Labrador Retriever (ROS_Cfam_1.0, GCF_014441545.1) with BWA‐MEM2 2.2.1. Alignments were coordinate‐sorted (SAMtools v1.13) and PCR/optical duplicates were marked with Picard MarkDuplicates (v2.21.8). After base quality score recalibration (BQSR), variants were called per sample with GATK v4.2.6.1 HaplotypeCaller (GVCF mode), jointly genotyped across samples (GenomicsDBImport → GenotypeGVCFs), filtered (VQSR/hard filtering) and annotated with snpEff (v5.0e). 99.91% of the sequenced reads mapped to the reference and 98.78% were properly paired, with a duplication rate of 5.70% and a mean mismatch rate of 0.63%. The deduplicated alignments covered 98.27% of the reference at ≥ 1× and showed a mean coverage of 29.3×. The genome was deposited at the European Nucleotide Archive with the accession number SAMEA10833662.
[内容请查看原文]
Variant Filtering
Starting from the multi‐sample, snpEff‐annotated cohort VCF comprising the affected dog and 1538 publicly available canine whole‐genome sequences (Table S1), snpEff annotations were flattened to one transcript effect per line, producing a tab‐delimited table of variant coordinates, alleles, per‐sample genotypes, and predicted functional consequences. As the mode of inheritance was unknown, both dominant and recessive disease models were considered. Candidate variants private to the affected dog were extracted using a custom pipeline. Variants were retained only where the case was heterozygous (0/1) or homozygous alternate (1/1) and all remaining animals were homozygous reference (0/0) or missing (./.). Private variants were subsequently filtered by retaining only protein‐changing variants, which were the variants with a SnpEff predicted impact of HIGH or MODERATE. Finally, the remaining variants were prioritized based on their presence in 16 known epidermolysis bullosa candidate genes (Has et al. 2020).
[内容请查看原文]
Variant Validation and Targeted Genotyping
The candidate variant LAMA3:XM_038543644.1:c.5690_5691ins240 was validated by Sanger sequencing and genotyped in all affected dogs, the available family members and one unrelated control dog by fragment length analysis. PCR amplification was performed using the QIAGEN Multiplex PCR Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions, with an annealing temperature of 60°C and 35 amplification cycles. The forward primer was 5′‐TCA TGG ATT AAA AAT GGA TGG TC‐3′ and the reverse primer was 5′‐TGC ATG GTT TCA AAT TCA CG‐3′. PCR products were treated with exonuclease I and alkaline phosphatase prior to fragment length analysis on a 5200 Fragment Analyzer (Agilent, Basel, Switzerland). Purified PCR products were additionally sequenced on an ABI 3730 DNA Analyzer (Thermo Fisher Scientific, Waltham, MA, USA) to determine the exact sequence of the inserted SINE element. Sequence chromatograms were analyzed using Sequencher 5.1 software (GeneCodes, Ann Arbor, MI, USA).
[内容请查看原文]
Results
[内容请查看原文]
Clinical and Histopathological Examination
Two female and one male random‐bred puppies were investigated. Two of the affected dogs were littermates, whereas the exact relationship to the third puppy was unknown. All three affected puppies shared similar clinical signs, which started at around 6–8 weeks of age and included complete loss of nails, deep ulcers in the paw pads and oral mucosa, vesicles and ulcers in the inner pinna, ulcers and bullae in the inguinal area and abdomen, and enamel hypoplasia (Figure 1a–c). Two of the puppies also suffered from unilateral non‐healing very large corneal ulcers (Figure 1d). All three affected puppies were euthanized a few weeks after diagnosis due to severe clinical deterioration, loss of appetite, and unmanageable pain.
[图：Clinical phenotype. (a) Multifocal ulcers in the abdominal area of an affected puppy. (b) Severe, deep paw‐pad ulcers. (]
Diagnosis of EB was confirmed by biopsies taken from lesional skin. The most consistent finding was dermo‐epidermal separation. It was accompanied by a range of reactive changes, which depended on the duration of the lesion and the presence of secondary bacterial infection. Lesions with different microscopic features were found near each other in most samples. In acute lesions, the epidermis was either of normal height or moderately hyperplastic and there was often accumulation of fibrin and/or blood in the space created by dermo‐epidermal separation (Figure 2a). Changes in the dermis consisted of vascular dilation, hypertrophy of fibroblasts and vascular cells, and mild neutrophilic infiltration. In subacute lesions, moderately hyperplastic epidermis was separated from dermis in which there was mild superficial fibrosis (Figure 2b). In longer‐standing lesions with evidence of secondary bacterial infection and/or recurrence, epidermal hyperplasia, dermal fibrosis and inflammation were more pronounced and there was ulceration, serocellular crusting, and epidermal regeneration (Figure 2c). Occasionally, neutrophils collected below the lifted epidermis (Figure 2d).
[图：Histopathological findings in skin biopsies from affected puppies. (a) Acute lesion. The epidermis is detached from the ]
[内容请查看原文]
Genetic Analysis
Comparison of the WGS data of the affected dog with 1538 publicly available canine genomes identified 214 heterozygous and 60 homozygous private protein‐changing variants (Table S2). Screening of 16 known EB candidate genes (Has et al. 2020), identified a single candidate variant in LAMA3. Although initially called as heterozygous, visual inspection of the short‐read alignments established that the variant was in fact a homozygous exonic SINE insertion into the LAMA3 gene (XM_038543644.1:c.5690_5691ins240; Figure 3a,b).
[图：Details of the LAMA3 variant. (a) Schematic representation of the genomic organization of the LAMA3 gene showing introns]
Sanger sequencing confirmed a 240‐bp SINE insertion in exon 45 of LAMA3 and determined its exact sequence and insertion site (File S1). The insertion was flanked by a 16‐bp target site duplication and contained a poly(A) tail. These two features were also evident from the short‐read alignments (Figure 3b and File S1). Sequence analysis showed that the insertion is in‐frame and predicted to introduce an additional 80 amino acids within the coiled‐coil domain of laminin α3 without introducing a premature termination codon (XP_038399572.1:p.(Lys1897_Glu1898insXaa[80])).
Fragment length analysis demonstrated complete concordance between genotype and phenotype in the investigated family. All three affected puppies were homozygous for the insertion, whereas their unaffected dam and the unaffected littermate were heterozygous carriers. The unrelated control dog was homozygous for the reference allele (Figure 3c,d).
[内容请查看原文]
Discussion
In this study, we identified a novel homozygous exonic SINE insertion in LAMA3 as the likely cause of junctional epidermolysis bullosa (JEB) in three random‐bred puppies. The variant was absent from 1538 publicly available canine whole‐genome sequences and represented the only protein‐changing private variant in a known EB candidate gene identified in the WGS data of a JEB affected dog. Genotypes at the variant co‐segregated with the disease phenotype within the investigated family. Applying the ACMG/AMP criteria for the interpretation of sequence variants in human diagnostics (Richards et al. 2015) or the AVCG v2 criteria (Desmet et al. 2026), we classified the variant as pathogenic (Table S3).
LAMA3 is a well‐studied candidate gene for JEB both in humans and animals (Capt et al. 2005; Graves et al. 2009; Has et al. 2020; Herrmann et al. 2021; Kivirikko et al. 1995; Sartelet et al. 2015). It encodes the laminin alpha 3 chain, one of the three subunits of laminin‐332, together with the β3 and γ2 chains encoded by LAMB3 and LAMC2, respectively. Laminin‐332 is a crucial component of the basement membrane that anchors epithelial cells to the underlying tissues. Its assembly begins in the endoplasmic reticulum, where the β3 and γ2 chains first form a heterodimer before incorporation of the α3 chain completes the mature heterotrimer. Consequently, the α3 chain is essential for efficient laminin‐332 assembly and secretion, and pathogenic variants affecting any of the three subunits disrupt the integrity of the complex and compromise dermal–epidermal adhesion (Bardhan et al. 2020; Parzymies 2026; Sugawara et al. 2008).
Assuming that the SINE insertion does not alter pre‐mRNA splicing, which was not investigated in the present study, the insertion is predicted to maintain the open reading frame and to introduce an additional 80 amino acids into the coiled‐coil domain of laminin α3. The predicted insertion contains a stretch of 15 phenylalanines encoded by the poly(A) tail of the SINE insertion. Such an amino acid sequence would clearly not be compatible with the heptad repeat organization of a coiled‐coil domain (Lupas 1996). The coiled‐coil domain mediates heterotrimer formation with the β3 and γ2 chains and is therefore critical for normal laminin‐332 assembly (Domogatskaya et al. 2012). Although the functional consequences of the insertion remain to be determined, the substantial in‐frame expansion within this highly conserved structural domain is expected to interfere with normal coiled‐coil formation and thereby severely compromise laminin‐332 assembly and/or secretion.
The predicted functional consequence of the identified variant is consistent with the severe clinical phenotype observed in the affected puppies. In humans, severe generalized JEB is typically associated with biallelic loss‐of‐function variants in LAMA3, LAMB3, or LAMC2, whereas variants permitting residual laminin‐332 expression generally result in milder phenotypes (Parzymies 2026). Similarly, Lama3 −/− null mice develop extensive skin blistering shortly after birth, exhibit abnormal hemidesmosomes, loss of anchorage function in the epidermis, and die within a few days of birth (Abrass et al. 2006; Yao 2017).
An interesting technical finding was that the causal variant was initially reported by GATK HaplotypeCaller as a heterozygous 53 bp insertion rather than a homozygous 240 bp SINE insertion. Visual inspection of the short‐read alignments revealed numerous soft‐clipped reads and discordant read pairs spanning the breakpoint, indicating the presence of a substantially larger insertion. As SINE elements are present in many copies throughout the canine genome, reads originating within the inserted element cannot be mapped uniquely and therefore receive low mapping‐quality scores, whereas reads spanning the unique flanking sequence align with high confidence. Filtering of low‐quality reads by the variant caller resulted in an artificial allelic imbalance, leading to an erroneous heterozygous genotype call. This observation illustrates a limitation of standard short‐read variant‐calling pipelines for repetitive element insertions and highlights the importance of visual inspection of candidate variants, particularly when evaluating plausible disease‐causing structural variants.
Although random‐bred dogs are often assumed to be genetically diverse, they can have high levels of inbreeding due to close matings such as between siblings or between parents and offspring (Leeb et al. 2023). The observed segregation pattern in this study is consistent with autosomal recessive inheritance and illustrates that inherited genetic disorders also occur in such populations, highlighting the importance of considering genetic diseases also in non‐pedigreed animals. It further expands the spectrum of pathogenic LAMA3 variants causing canine JEB and demonstrates the value of whole‐genome sequencing combined with careful interpretation of structural variants for the diagnosis of inherited skin disorders in veterinary patients.
Author Contributions
Sarah Kiener: investigation, visualization, writing – original draft, writing – review and editing. Ronnie Kaufmann: conceptualization, investigation, visualization, writing – original draft, writing – review and editing. Tosso Leeb: conceptualization, visualization, writing – original draft, writing – review and editing, funding acquisition. Ori Brenner: conceptualization, investigation, visualization, writing – original draft, writing – review and editing. Vidhya Jagannathan: data curation, writing – review and editing.
Funding
This research was funded by the Swiss National Science Foundation (Grant 310030_200354).
Ethics Statement
All examinations and animal experiments were carried out after obtaining written informed owner's consent and in accordance with local laws, regulations, and ethical guidelines. Dog blood samples were collected with the approval of the Cantonal Committee for Animal Experiments (Canton of Bern, Switzerland; permits BE 71/19 & BE94/2022). Skin punch biopsies were taken for diagnostic purposes with informed consent of the owners, in accordance with ethical guidelines and approved procedures.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
[内容请查看原文]
[内容请查看原文]
[内容请查看原文]
[内容请查看原文]