A homozygous lamin B receptor variant resulting in Pelger–Huët anomaly without skeletal dysplasia

ABSTRACT

The lamin B receptor (LBR) is an inner nuclear membrane protein with a structural function affecting nuclear segmentation in neutrophils and an enzymatic function as a sterol reductase. Heterozygous LBR variants cause nuclear hyposegmentation (Pelger–Huët anomaly), while homozygous LBR pathogenic variants cause a spectrum of systemic malformations up to severe skeletal deformities and prenatal death. Already in 1981, a homozygous Pelger–Huët proband with ovoid neutrophil nuclei but without special clinical symptoms has been reported. Later it was shown that the amount of LBR was reduced, perhaps due to a regulatory variant. We have now re-analyzed the proband´s DNA and confirmed the existence of a homozygous missense variant (NM_002296.4): c.1445T > G; p.V482G. However, further molecular testing did not reveal any regulatory variant. The missense variant was further investigated: It affects an evolutionary highly conserved amino acid, which is in direct contact with the inner nuclear membrane, but has no direct influence on the LBR sterol reductase activity. A homozygous valine to glycine substitution has not been detected in any study so far. Therefore, we conclude that this variant has a significant effect on the structural function of LBR but no clinically relevant effect on its enzymatic function.

Introduction

The nucleus of almost all human cells is ovoid. An exception is the highly deformable, lobulated nucleus of neutrophils, allowing for rapid diapedesis through blood vessel walls and migration to the sites of bacterial or fungal infections. It is an essential component of the innate immune system, an evolutionary ancient trait. The flexibility of the neutrophil nucleus is particularly due to the paucity of lamins A/C and B1 and the elevated levels of the lamin B receptor (LBR). LBR is an inner nuclear membrane protein with a structural function interacting with chromatin and lamins, and an enzymatic function as a sterol reductase [1–6]. Heterozygous LBR variants cause nuclear hyposegmentation in neutrophils (Pelger–Huët anomaly, PHA); the carriers are otherwise asymptomatic [7]. However, homozygous LBR alterations published so far cause a spectrum of systemic malformations ranging from cardiac defects, brachydactyly, and intellectual disability as occurs in homozygous PHA, to severe skeletal deformities and prenatal death, as found in Greenberg dysplasia [7–10]. This combination of an enzymopathy and a structural trait in the same protein is quite unique. Interestingly, the father of a compound heterozygous fetus with Greenberg dysplasia had normal lobulated neutrophils and an LBR missense variant, p.R583Q. Consequently, this pathogenic variant affects the sterol reductase domain only and thus uncouples the metabolic from the structural function of LBR [11].

Already in 1981, Aznar and Vaya reported an 18-month-old girl with clear Pelger–Huët unsegmented, ovoid neutrophil nuclei but without any special symptoms and normal intelligence, except polydactyly of hands (six fingers on one hand) and feet (six toes on both feet) [12]. Polydactyly was also observed in the proband´s sister, who had normally lobed granulocyte nuclei. Thus, these symptoms are obviously unrelated to the Pelger trait. This proband was later shown to carry a homozygous interval including LBR, and later, a reduced amount of LBR in lymphoblastoid cells was detected [13]. The DNA analysis using Sanger sequencing was performed in 2010 and revealed a missense variant (NM_002296.4): c.1445T > G; p.V482G which was not regarded as causative at this time, and a regulatory variant was suspected. In the present study, we further investigated the proband´s DNA and RNA by state‑of-the‑art genomic technologies with the aim of identifying the potential regulatory variant in LBR that is causative for the Pelger–Huët phenotype.

Materials and methods

Subjects

We obtained blood smears from a homozygous Spanish woman with the Pelger–Huët anomaly (see [12] for details) and peripheral blood to establish a lymphoblastoid cell line. The clinical and molecular data of the homozygous, heterozygous, and control individuals involved in this study have already been published as well as the data of the other homozygous patient with a splice-site variant, a deletion of six bp in intron 12 (IVS12-5-10del), and his heterozygous parents [13,14].

Characterization of nuclear segmentation in neutrophils

Two investigators independently analyzed the nuclear segmentation on coded slides, based, among others, on the ‘rule of thirds.’ This analysis has already been reported [13].

Protein analysis

Lymphoblastoid cells in the logarithmic growth phase were counted, washed, and proteins extracted (see [1] for details). For western blotting the samples were adjusted to identical protein concentrations and loaded on a 4–12% Bis-Tris-Gel (Invitrogen). Gel electrophoresis was performed under denaturing conditions using an XCell Blot Module (Invitrogen). Guinea pig anti-LBR serum was used at 1:500 dilution; an antibody against β-actin confirmed equal protein loading, and the species-specific secondary antibody was Cy3-goat anti-guinea pig (Dianova, Hamburg, Germany). The results of this analysis have already been reported [13].

Genetic analysis

Initial variant screening in LBR was performed by Sanger sequencing of PCR products on a MegaBACE 1000 sequence detection system using DYEnamic ET Terminator Cycle Sequencing Kit (Amersham Pharmacia Biotech, see [1] for details). To detect further potentional regulatory variants in LBR, Illumina-based PCR-free genome sequencing was performed. Library preparation was performed to the manufacture´s instruction. Sequencing was performed on a NovaSeq X Plus System (Illumina) with a mean coverage of 30×. Sequence reads were mapped to the genome version GRCh37 (UCSC hg19) with the Burrows-Wheeler Aligner (BWA MEM) [15]. Single-nucleotide variants and short indels were called with the Genome Analysis Toolkit (GATK) according to the GATK Best Practices. We used Mehari (publication in preparation) for variant annotation, and Varfish for filtering and further data analysis as described previously [16,17].

In addition, we performed a poly-A enrichment from total RNA preparations from an immortalized lymphoblast culture of the proband. We performed transcriptome sequencing to investigate aberrant LBR splicing. Sequencing was performed as previously described [16]. In short: After poly-A selection, mRNA was purified and fragmented, and the fragments underwent reverse transcription using random primers. This was followed by second‑strand cDNA synthesis, and after end repair and A-tailing, indexing adapters were ligated. The purified products were amplified (15 PCR cycles) to create the final cDNA library. This was sequenced on an Illumina NovaSeq6000 sequencing instrument (Illumina, SanDiego, CA, USA) with a PE100 protocol. Sixty to seventy million sequence reads were generated. The RNA-Seq reads were mapped to the human genome (GRCh37, Ensemble release 110) with STAR version-2.7.11a43. The LBR transcript was visualized using the Integratve Genomics Viewer (IGV), and rare variants, identified using genome sequencing, were inspected.

Moreover, an exploratory differential expression analysis (DESeq2 applied to Salmon-quantified reads) of LBR and seven housekeeping genes (RPL13A, SDHA, UBE2D2, IPO8, HPRT1, PPIA and MT-APT6) was performed in the lymphoblastoid cell line of the LBR proband and four lymphoblastoid control cell lines [18–20]. Furthermore, we investigated the expression of LMNB1, LMNB2 and LMNA following the same approach.

Results and discussion

The first molecular analysis of the homozygous Pelger–Huët proband was performed in 2010 [13]. At that time, we characterized the nuclear segmentation in neutrophils of the proband and the amount of LBR protein in her lymphoblastoid cells in comparison with homo- and heterozygotes for a splice-site variant, a deletion in intron 12 (IVS12-5-10del), and a homozygote for the wild-type allele. The results are summarized in Figure 1. Clearly, the homozygous proband had a lower lobulation index than the heterozygotes. This points toward a significant alteration interfering with the structural role of LBR. In addition, the amount of LBR protein was reduced, but obviously high enough that its functional role as sterol reductase is not affected and the patient shows no further clinical features besides the Pelger trait. Linkage analysis of the proband and her mother revealed a homozygous interval on chromosome 1q42 in the proband and a homozygous missense variant in exon 11 in the LBR gene, which is located within this interval (Suppl. Table 1). This variant was classified as ‘benign’ at the time of analysis leading to the suspicion of an unknown regulatory variant as an explanation of the Pelger phenotype. To solve this riddle, we have now sequenced the proband’s DNA again using short-read genome sequencing to investigate whether regulatory variants can be detected.

Figure 1.

Effects of 0 to 2 wild‑type copies of the LBR-gene on nuclear segmentation and LBR protein level. a) Nuclear morphology in neutrophils of the proband, homozygous for the missense mutation V482G, and an individual homozygous for a splice-site mutation in intron 12 (IVS12-5-10del), his heterozygous parents, and a proband homozygous for the wild‑type allele. Scale bar 10 μm. [after 13] b) Nuclear segmentation determined according to the rule of thirds on 50 neutrophils each (controls 600 cells from 12 individuals). [after 14] c) Amount of LBR protein in western blots from lymphoblastoid cells of the 5 probands. [after 13].

Using Varfish, we investigated the homozygous region on chromosome 1, including LBR: 225,588,814 to 225,615,831 (GRCh37 coordinates). Filtering for rare homozygous variants within this region revealed two variants, including the already known alteration LBR (NM_002296.4): c.1445T > G; p.V482G (Suppl. Figure 1). In addition, another deep intronic variant was detected, LBR (NM_002296.4): c.367-978A > G (Figure 2).

Figure 2.

Visualization of short-read genome sequencing and RNA-seq data from lymphoblastoid cells of the proband. Each panel includes tracks for total coverage, junction coverage, predicted transcripts, and read alignments. GS: reads of genomic sequences including those of the introns. RNA-seq: RNA-sequences that span the exons are depicted as arches. First panel: RNA-seq read alignments across the complete LBR gene; second panel: RNA-seq read alignments across exons 11 (with the homozygous missense variant) and 12; third panel: illustrating that the homozygous intronic variant between exons 3 and 4 has no effect on the canonical splicing pattern of the LBR transcript (NM_002296.4).

The homozygous variant c.1445T > G; p.V482G is extremely rare and has been found in only one individual in a heterozygous state in gnomAD v4.1.0. In silico prediction of this variant revealed the following pathogenicity scores: PolyPhen: 0.317 (benign), CADD: 23.9 (less deleterious), AlphaMissense: 0.33 (likely benign), phyloP: 6.61 (more conserved sequence), REVEL:0.554 (neutral) and Primate-AI-3D: 0.78 (possibly damaging). Most of these parameters indicate no impact or a mild impact on protein function. The second homozygous variant in LBR, c.367-978A > G, is a deep intronic alteration not detected in gnomAD v4.1.0. In silico prediction for such deep intronic variants is limited, and therefore, the interpretation of this nucleotide change remains problematic; the predictions, especially in terms of cryptic splice sites, are limited. Therefore, and because of the reduction of LBR abundance, we decided to perform transcriptome sequencing to investigate whether we can detect aberrant splicing. We inspected the LBR locus in general and analyzed the two variant positions in detail using IGV. We could not detect any changes in the canonical splicing pattern of the LBR transcript (NM_002296.4) in the proband´s lymphoblastoid cells (Figure 2).

Due to this extensive testing and the exclusion of a regulatory variant, the formally known missense variant must be considered as the cause of the Pelger trait in this family. We therefore investigated the properties of this variant in more detail. The evolutionary comparison also shows that valine or the equally nonpolar and neutral amino acids leucine or isoleucine are predominantly found, indicating that these properties are important at this position (Figure 3(A)). In addition, we observed that within the known secondary structure of LBR the p.V482G variant maps to the transmembrane domain 7 and is in direct contact with the inner nuclear membrane (Figure 3(B)). In contrast to valine, leucine, and isoleucine, glycine has no side chain. It is therefore possible that this change at this particular position might affect nuclear segmentation.

Figure 3.

Properties and localization of the LBR variant p.V482G. A. Evolutionary conservation of the original amino acid (Valine 482). [from https://media.githubusercontent.com/media/aminodektc/42/master/LBR/LBR.png]. B. Schematic view of the lamin B receptor as a protein of the inner nuclear membrane. Shown is a section of domains 7 and 8 with variants at positions 482 and 583, which preferentially affect nuclear morphology (Pelger anomaly) or enzyme function (Greenberg dysplasia), respectively. [after ²¹]. C. 3D protein structure of LBR in AlphaFold. The arrow points to amino acid Val482, the position of the mutation, both in the overall structure of the protein and its enlarged view. The model confidence is color coded; dark blue indicates very confident (pLDDT > 90), light blue confident (90 > pLDDT > 70), yellow low confidence (70 > pLDDT > 50), orange is very low confidence (pLDDT < 50). The NADPH binding pocket includes the amino acids N547 and R583, important for LBR´s sterol reductase activity [7, 25] D. Location of the 10 homozygous LBR missense variants detected so far [after 5, 27], resulting in Greenberg dysplasia (red), Pelger–Huët anomaly with skeletal anomalies (black), and the present case with Pelger–Huët anomaly without any anomaly (blue). The transmembrane domain (TM) 6 is amino acid 447–467, TM7 481–501, and TM8 561–581. [9,18] Thus, the homozygous missense variants L456V and D460A are in TM6, and V482G is in TM7; the other variants at amino acid positions 512, 547, 558, 583, and 586 are in the connecting regions.

Clearly, the reduced LBR amount could also affect nuclear morphology. As the amount of LBR was only analyzed in lymphoblastoid cells, its exact value during granulopoiesis is unknown. Independent of this, the variant likely impairs the structural function of LBR, resulting in altered nuclear segmentation.

According to AlphaFold [22,23], the LBR amino acid position 482 is not part of the NADPH binding pocket. Its influence on the LBR sterol reductase activity is obviously not high enough to result in clinically relevant consequences (Figure 3(C)). However, the high-confidence structural prediction of the LBR proteins by AlphaFold also has its limitation [24] as it could not predict the interaction of p.482 with the inner nuclear membrane.

Next, we investigated all diseases caused by homozygous missense variants known so far mapping to the sterol reductase domain. These variants caused a phenotypic spectrum from PHA with skeletal anomalies to embryonic lethal Greenberg dysplasia (Figure (3D)). The homozygous missense variants L456V and D460A are in the transmembrane domain 6 and V482G in transmembrane domain 7; the other variants at amino acid positions 512, 547, 558, 583, and 586 are in the connecting regions. The only exception is the current case, in which no skeletal dysplasia has been detected. This also applies when all known homozygous and compound heterozygous LBR variants are considered (Suppl. Table 2).

In strong contrast to p.V482G, multiple computational pathogenicity prediction scores indicate a deleterious effect for the missense variant p.R583Q, identified in a severely affected stillborn with Greenberg dysplasia whose heterozygous father had normal neutrophils with multisegmented nuclei: PolyPhen: 0.999 (probably damaging), CADD: 32 (rather deleterious), AlphaMissense: 0.857 (likely pathogenic), phyloP: 6.42 (more conserved sequence), REVEL: 0.947 (pathogenic supporting) and Primate-AI-3D: 0.85 (possibly damaging). This supports the idea that homozygosity for this variant might cause a Greenberg dysplasia phenotype. This variant was located within the sterol reductase domain (Figure 3(C)), potentially impairing this domain but leaving structural functions of LBR intact, in contrast to the variant p.V482G [11].

Interestingly, the variant p.V482G affects protein level, which is not uncommon for genetic variants [23]. Based on an exploratory differential expression analysis (DESeq2 applied to Salmon-quantified reads), we identified an increased mRNA expression of LBR compared to four control cell lines and in comparison with multiple housekeeping genes commonly used as reference genes [20] (Suppl. Figure 2). The lamins are not upregulated. These transcriptome data show that the LBR mRNA of the proband´s lymphoblastoid cell line is not reduced. Applied to granulocytes, this argues against a regulatory effect as an explanation for the reduced LBR protein abundance. Furthermore, these data point toward a compensatory upregulation of LBR, likely as a consequence of an increased protein instability. Independent of this, the data indicate that the remaining protein is sufficient to prevent the proband from clinical problems. Due to its structural role and its enzymatic activity, the genotype–phenotype correlation of LBR variants is a fascinating aspect covered in a number of recent articles and reviews [4–10,26,27]. Somewhat simplified, alterations can be categorized into those variants that, depending on the amount of the protein and the domain affected, preferentially affect the structural or enzymatic function of the protein. Moreover, mice with Lbr mutations (mouse ichthyosis locus) provide an ideal model for determination of the LBR function in normal and pathological states [28]. Interestingly, a mouse model with deletion of LBR N-terminal domains and intact sterol reductase domains also shows the PHA phenotype without skeletal defects [29].

To the best of our knowledge, the present homozygous variant p.V482G is the first human case in which no symptoms resulting from enzymatic deficiencies of LBR were observed, but a significant effect on neutrophil nuclear morphology, thus uncoupling the deficient structural from the metabolic function of LBR. It is, however, still open which role plays the altered LBR protein and/or its reduced amount in this context. Because of the bifunctional role of LBR, the characterization of the homozygous p.V482G variant as ‘benign’ or likely benign is correct with a view on the enzymatic function. However, concerning its structural effect the classification as ‘possibly damaging’ is more adequate. It should be noted that the new computational tools can predict the impact of variants on protein function more accurately than 15 years ago when we performed the first analysis of the proband´s LBR gene [13]. At that time the missense variant was classified as ‘benign,’ pointing to an unknown regulatory alteration. However, this interpretation in light of the observations in human and mice might be regarded as an example, where the in silico prediction is not conclusive in the scenario of a variant causative for a subclinical trait.

Funding Statement

This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG), grant number Sp 144/18–1 to K.S. and K.H. [SFB 577, project A4].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available in this article.

The additional datasets generated during the current study are available from the corresponding authors on reasonable request.

Ethics approval and consent to participate

This study was performed according to the Declaration of Helsinki principles of medical research involving human subjects and was approved by the institutional Ethics Committees of the Charité-Universitätsmedizin Berlin.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/19491034.2026.2720062