Abstract

   Mucopolysaccharidosis (MPS) encompasses a heterogeneous group of
   lysosomal storage diseases resulting from mutations in genes encoding
   lysosomal enzymes responsible for the degradation of
   mucopolysaccharides, also known as glycosaminoglycans (GAGs). Current
   therapeutic strategies for MPS include hematopoietic stem cell
   transplantation (HSCT), enzyme replacement therapy (ERT), and
   symptomatic therapy. This study investigated dynamic changes in MPS
   type II (MPS-II) through genomic and single-cell sequencing in a
   patient undergoing ERT. Analysis of peripheral blood mononuclear cells
   (PBMCs) from one MPS-II patient of 10 year old at different disease
   stages through scRNA-seq identified various immune cell types,
   including natural killer (NK) cells, NKT cells, CD4 + and CD8 + T
   cells, CD14 + and CD16 + monocytes, and B cells. Monocytes and
   macrophages were significantly reduced during the severe stage of
   MPS-II but increased during the recovery stage following ERT. Notably,
   monocyte subtype mono3 was exclusively expressed in the severe stage,
   while mono1_2, a subtype of mono1, was absent during the severe stage
   and exhibited distinct biological functions. These findings suggest
   that monocytes and macrophages play critical roles in the pathogenesis
   of MPS-II and in the response to ERT. Pseudotime, Gene Ontology, and
   cell-communication analyses revealed unique functions for the different
   cellular subtypes. Notably, key molecules mediating cellular
   interactions during ERT in MPS-II included CXCR3, PF4, APP, and C5AR1
   in macrophages, RPS19 in T cells, HLA-DPB1 in B cells, ADRB2 in NK
   cells, and IL1B, C5AR1, RPS19, and TNFSF13B in monocytes. Overall,
   integrative analysis delineated the expression dynamics of various cell
   types and identified mutations in MPS-II, providing a comprehensive
   atlas of transcriptional programs, cellular characterizations, and
   genomic variation profiles in MPS-II. This dataset, along with advanced
   integrative analysis, represents a valuable resource for the discovery
   of drug targets and the improvement of therapeutic strategies for
   MPS-II.

Supplementary Information

   The online version contains supplementary material available at
   10.1038/s41598-025-97330-7.

   Keywords: Type II mucopolysaccharidosis, Enzyme replacement therapy,
   Single-cell RNA sequencing, T-cell receptor sequencing, Whole-genome
   sequencing

   Subject terms: Endocrine system and metabolic diseases, Disease
   genetics

Introduction

   Mucopolysaccharidosis (MPS) comprises a heterogeneous group of
   lysosomal storage diseases caused by mutations in genes encoding
   lysosomal enzymes responsible for the degradation of
   mucopolysaccharides, also termed glycosaminoglycans (GAGs)^[36]1. There
   are seven recognized types of MPS, which include a total of 13
   subtypes, each defined by specific gene mutations that result in the
   dysfunction of corresponding enzymes responsible for GAG
   degradation^[37]2. MPS types can be categorized into somatic and
   neuronopathic forms. Type I MPS is caused by the deficiency of
   a-l-iduronidase, whose deficiency results in accumulation of
   GAGs^[38]3. Type III MPS (Sanfilippo syndrome) involves multiple enzyme
   deficiencies involving heparan-N-sulfatase (IIIA),
   a-N-acetylglucosaminidase (IIIB), a-glucosaminide acetyltransferase
   (IIIC), and N-acetylglucosamine-6-sulfatase (IIID), causing
   accumulation of HS. MPS IIIA and IIIB are associated with mutations in
   the genes SGSH and NAGLU, respectively^[39]4. Type IV MPS (Morquio
   syndrome) is of two types: type IVA which occurs due to the deficiency
   of N-acetylgalactosamine-6-sulfatas and Type IVB which occurs due to
   the deficiency of b-galactosidase enzyme^[40]3. Type VI MPS (Maroteaux
   Lamy) occurs due to the accumulation of DS and chondroitin-4-sulfate
   (C4S) as a result of deficiency of the enzyme
   N-acetylgalactosamine-4-sulfatase involving gene ARSB^[41]3. Partial
   degradation and accumulation of C4S, chondroitin-6-sulfate (C6S), DS,
   and HS as a result of b-galactosidase deficiency due mutation in the
   gene GUSB leads to Type VII MPS (Sly syndrome)^[42]3. Type IX MPS
   (Natowicz syndrome) is the rarest of all types and occurs due to the
   deficiency of hyaluronidase enzyme resulting in accumulation of
   hyaluronan^[43]3. MPS type II (MPS-II), also known as Hunter syndrome,
   is the only MPS with X-linked inheritance, caused by a deficiency in
   the enzyme iduronate-2-sulfatase (IDS), leading to the progressive
   accumulation of dermatan sulfate (DS) and heparan sulfate (HS)^[44]5.

   Current therapeutic strategies for MPS include hematopoietic stem cell
   transplantation (HSCT), enzyme replacement therapy (ERT), and
   symptomatic therapy^[45]6,[46]7. Several clinical trials of gene
   therapy for MPS-II were conducted, including retroviruses with genome
   editing systems such as zinc finger nucleases or CRISPR/Cas9^[47]8.
   Clinical trials of adeno-associated viruses were also conducted for
   MPS-II patients, which could improve central nervous system function of
   patients^[48]9,[49]10. ERT can reduce the accumulation of GAGs in
   cells, alleviating hepatosplenomegaly and improving cardiopulmonary
   function and joint mobility, but conventional ERT exhibits limited
   efficacy in improving cognitive and central nervous system functions
   due to the blood-brain barrier (BBB). New types of ERT using monoclonal
   antibody (MAb) that functions as a vehicle to bypass the BBB are under
   clinical trials for MPS-II^[50]11. Another method to penetrate the BBB
   is to change the injection route, via IT or intracerebroventricular;
   however, IT ERT for MPS-II did not demonstrate therapeutic efficacy and
   have been discontinued^[51]12. Although early HSCT can significantly
   improve neurological symptoms, the risk of graft-versus-host disease
   and long-term use of multiple immunosuppressants limit its
   applicability. HSCT can stabilize or prevent hydrocephalus and prevent
   the deterioration of psychomotor functions, and also has been
   associated with improvements in cognitive function and CNS
   manifestation^[52]13,[53]14. Previous meta-analysis has demonstrated
   that ERT in MPS-II reduces urinary GAG content and exhibits a
   dose-response relationship, while several animal model studies have
   reported IDS deficiency and increased GAG levels in the context of
   MPS-II^[54]15–[55]17. However, these studies failed to capture
   molecular heterogeneity, necessitating the observation and measurement
   of ERT effects at the transcriptome level.

   Single-cell RNA sequencing (scRNA-seq) enables the identification of
   cell subpopulations and their distinct functions by exploring the
   transcriptomic profiles of individual cells. This technology has
   emerged as a powerful tool for elucidating pathophysiological changes
   in diseases^[56]18 and profiling gene expression at the single-cell
   level^[57]19. Recently, scRNA-seq has revolutionized and expanded our
   ability to investigate critical questions in metabolic diseases, such
   as type 1 diabetes, adipogenesis, osteoclast function, and pancreatic
   beta-cell differentiation^[58]20–[59]22. Previous studies explored
   transcriptomic changes in MPS through RNA sequencing^[60]23. However,
   no research has focused on MPS-II to measure how cells change during
   treatment at the single-cell level.

   In this study, whole-genome sequencing (WGS) was performed on an MPS-II
   patient, along with scRNA-seq, T-cell receptor sequencing (TCR-seq),
   and B-cell receptor sequencing (BCR-seq) at two different ERT stages.
   This comprehensive characterization of cell dynamics was conducted
   using peripheral blood mononuclear cells (PBMCs). The WGS results
   identified 16 IDS germline mutations in the MPS-II patient, with only
   one frame shift deletion mutation remaining after filtering. The
   scRNA-seq results revealed that monocytes and macrophages almost
   disappeared in severe MPS-II and increased during the recovery stage
   following treatment. This suggests that monocytes and macrophages play
   crucial roles in MPS-II pathogenesis and response to ERT. Our genomic
   and single-cell transcriptomic analyses provide an atlas of
   transcriptional programs, cellular characterizations, and genomic
   variation profiles in MPS-II, representing a valuable resource for
   MPS-II drug target discovery and therapeutic strategy improvement.

Methods

Patient material, ethics statement, and consent for publication

   MPS was diagnosed based on clinical manifestations,
   iduronate-2-sulfatase deficiency, and IDS gene mutations. The MPS
   patient received four treatments with iduronate-2-sulfatase at a dosage
   of 0.5 mg/kg. PBMCs were collected after the first and second treatment
   stages. Clinical data for the patient are provided in Table [61]1. All
   procedures involving human subjects and tissues were approved by the
   Ethics Committee of the Children’s Hospital of Chongqing Medical
   University (NO.202364), with signed informed consent obtained from the
   patient’s parents.

Table 1.

   Clinical information of MPS II patient.
   Basic Information
   Sex                     Male
   Onset age               2 years old
   Diagnosis age           10 years old 2 months
   Diagnosis date          January 2022
   Clinical manifestation

   Coarse faces: macrocephaly, prominent forehead, thick hair, blunt nasal
   tip, and thick lips

   Dysostosis multiplex: claw hands, joint stiffness/contracture/
   enlargement, short stature

   Nervous system: intelligence Regression, behavioral abnormalities,
   disturbed sleep, epilepsy, IQ/DQ assessment was not completed,
   FMFM98points (98/183, 66.42%), GMFM203points (203/260, 78.00%)

   Ear-nose-throat and mouth: airway obstruction, tonsil/adenoid
   hypertrophy, snoring, recurrent respiratory infections

   Cardiovascular system: the aortic valves regurgitation, ventricular
   hypertrophy and heart failure

   Digestive system: abdominal distension, evident hepatosplenomegaly
   UGAGs (DMB)             +
   IDS(DBS)                0.00(>1.500 nmol/h/mL)
   WGS                     IDS (c.1033delT)
   Clinical classification Severe
   ERT start time          April 7th, 2022
   ERT start time          April 28th, 2022
   MPS1: before ERT MPS2: after ERT
   Assessment age 10-years-old 2months 10-years-old 6months
   Clinical manifestation
    Respiratory infection 2–3/month 1/month
    Seizures Up to 7 times a day Absent
    Sleep quality Unable to sleep on back, snore, Occasional apnea, sleep
   light Able to sleep on back for 3–4 h
    6MWT Walk 6 m alone, easy to fall Walk 20 m alone
    3MSC Unable to climb stair alone Able to climb 5 stairs alone
    Volume of liver and spleen Liver was 5 cm below the costal margin,
   spleen was 3 cm below the costal margin Liver was 4 cm below the costal
   margin, below spleen was 3 cm the costal margin
    Height 116.5 cm 119.5 cm
   [62]Open in a new tab

   FMFM : Fine Motor Function Measure.

   GMFM : Gross Motor Function Measure.

   UGAGs: Urinary glycosaminoglycans.

   DMB: Simethyl methylene blue.

   IDS: Iduronate-2-sulfatase.

   DBS: Dried blood spot.

   WGS: Whole Genome Sequencing.

   ERT: Enzyme replacement therapy (idursulfase-β, Hunterase^®).

   MWT: Minute walking test.

   MSC: Minute stair climbing experiment.

Isolation of PBMCs, single-cell library construction, and VDJ-sequencing

   Ficoll-Paque PREMIUM was used to isolate PBMCs from 3 mL of
   EDTA-treated blood from the patient following the manufacturer’s
   manual. For scRNA-seq data, patient cell suspensions were pooled in
   equal volumes and diluted to a concentration of 1 × 10^6 cells/mL.
   There was a slight donor bias in the number of cells from each sample
   in the mixture. The mixed suspension was then loaded onto the 10×
   Genomics platform to create a barcoded cDNA library for individual
   cells. The individual libraries were pooled and sequenced using the
   Illumina HiSeq 2500 platform. Cryopreserved samples were thawed at 37 ℃
   and washed twice in ice-cold 1 × phosphate-buffered saline (PBS).
   ScRNA-seq, TCR-seq, and BCR-seq were performed on CD138 + and
   CD138 − bone marrow mononuclear cells (BMMCs) using the Chromium Next
   GEM Single-cell 5’ Reagent Kit v1.1 and V(D)J Reagent Kit v1.1
   according to the protocols provided by the manufacturer (14 000
   cells/channel). Generated gene expression libraries were paired-end
   sequenced using the NovaSeq 6000 S2 platform and generated V(D)J
   libraries were paired-end sequenced using the NextSeq500 platform.

ScRNA-seq

   Freshly prepared PBMC suspensions were processed immediately using a
   10× Chromium 3’ v3 Kit (10× Genomics, Pleasanton, CA, USA), in
   accordance with the manufacturer’s protocols. The libraries were
   sequenced using the Illumina NovaSeq 6000 platform with paired-ended
   sequencing and dual indexing. Sequencing involved 26 cycles for Read 1,
   eight cycles for the i7 index, and 98 cycles for Read 2. On average,
   each sample generated approximately 709 million reads (standard
   deviation (SD) of 121 million).

Data processing, integration, and cluster annotation

   Raw BCL files were demultiplexed and mapped to the reference genome
   GRCh38 using Cell Ranger Single-cell (v1.1.0). Doublets were removed
   through DoubletFinder (v2.0.3). The filtered count matrix was converted
   to a Seurat object using Seurat (v4.3.0). Cells with more than 3 000
   genes, fewer than 200 genes, or more than 15% mitochondrial counts were
   excluded. Data normalization for each sample was performed using the
   NormalizeData method, and the top 3 000 highly variable genes were
   selected for subsequent dimensionality reduction. Batch effects among
   samples were corrected using the integration method in Seurat.
   Principal component analysis (PCA) and uniform manifold approximation
   and projection (UMAP) were used to reduce dimensionality and visualize
   cell distribution. Cluster-specific genes were identified using the
   Wilcoxon rank-sum test in the FindAllMarkers function (fold-change ≥ 2
   and adjusted p ≤ 0.05). Cell clusters were annotated based on
   well-recognized marker genes.

Cell-cell communication analysis

   To investigate the molecular interaction networks between different
   cell types, CellPhoneDB (v3.0.0)^[63]24 was used to infer cell-cell
   communication from the combined expression of multi-subunit
   ligand-receptor complexes. Ligand-receptor pairs (P < 0.05) were
   retained to assess the relationships among different cell clusters.

Single-cell trajectory and RNA velocity analyses

   R package Monocle2^[64]25 was used to order cells in pseudotime along a
   developmental trajectory. After dimensionality reduction, cells were
   arranged in pseudotime to reflect their progression through the
   developmental program, enabling single-cell trajectory analysis of cell
   subtypes using Monocle2. Velocyto R package (v0.6)^[65]26 was used to
   evaluate cell lineage dynamics and RNA velocity. Spliced and unspliced
   count matrices were constructed using the Velocyto pipeline based on
   “Cell Ranger” output files. Loom files for each sample were created and
   merged, retaining the top 3 000 most variable genes, with a minimum of
   100 spliced counts. After data normalization and scaling, cell clusters
   were embedded with velocity streams based on the ratio of spliced to
   unspliced counts and projected onto a UMAP plot for visualization.

Pathway and gene ontology (GO) enrichment analysis

   Gene set enrichment analysis (GSEA) was performed using ClusterProfiler
   (v4.6.2)^[66]27. Differentially expressed genes (DEGs) (adjusted
   P < 0.05 and |Log2fold-change (FC) > 0.25) from each PBMC subpopulation
   were used for Kyoto Encyclopedia of Genes and Genomes (KEGG)^[67]28 and
   GO enrichment analysis. Pathways and biological processes with P < 0.05
   were considered significantly enriched.

TCR/BCR-seq and copy number variation (CNV) analyses

   VDJ profiled_contig_annotations files were loaded into R, combined for
   all patients (separately for TCR-seq and BCR-seq), and mapped to the
   Seurat object using the scRepertoire package (v1.8.0). CNV features of
   cell clusters were examined using CopyKat to distinguish aneuploid
   cells in each sample with default parameters^[68]29.

WGS and mutation calling

   Genomic DNA was extracted from frozen blood plasma samples, with
   library preparation details provided in the Supplementary Materials and
   Methods. The DNA libraries were sequenced using the Illumina HiSeq
   platform, generating 150-bp paired-end reads. The raw sequencing data
   were mapped to human reference genome hg38 (GATK Resource Bundle) using
   BWA-MEM (v.0.7.17)^[69]30. Following GATK best practices, the .bam file
   was sorted and indexed, with polymerase chain reaction (PCR) duplicates
   marked and removed and base quality recalibrated. Germline mutations
   were detected using GATK HaplotypeCaller^[70]31, Strelka^[71]32, and
   VarScan^[72]33, retaining only those mutations called by at least two
   algorithms. Bam-readcount^[73]34 was used to quantify the number of
   reference and alternative alleles, filtering variants with fewer than
   five alternative allele counts or an alternative allele frequency below
   20%. Mutations were also filtered for rare variants with an allele
   frequency below 0.05% in the 1000 Genomes and Genome Aggregation
   Database (gnomAD).

Results

Single-cell transcriptome and genome landscape of PBMCs in an MPS-II patient
during ERT

   The MPS-II-diagnosed patient exhibited significant clinical features,
   including typical facial characteristics such as alar hypertrophy,
   enlarged tongue, thick lips, and clawed hands. Imaging studies showed a
   J-shaped sella on X-ray and an enlarged supratentorial region on MRI
   (Fig. [74]1A). Other clinical manifestations were provided in
   Table [75]1. To investigate the composition and diversity of PBMCs at
   different stages of clinical treatment, scRNA-seq, TCR-seq, and BCR-seq
   were performed on PBMCs from an MPS-II patient at two ERT stages (MPS
   II (a): severe stage before treatment; MPS II (b): recovery stage after
   treatment), and related clinical performance was summarized in
   Table [76]1. Combined with a public scRNA-seq dataset of PBMCs from
   three healthy children^[77]35, a total of 56 479 cells were obtained
   with an average of 2 000 genes per cell after quality control. Cell
   clusters were visualized using UMAP and annotated into nine cell types,
   including B cells, CD14 + and CD16 + monocytes, CD4 + and CD8 + T
   cells, macrophages, NK cells, NKT cells, and erythrocytes (FigS. [78]1B
   and C, S1-A). Analysis of cell type counts and composition in each
   sample indicated a marked reduction of monocytes and a limited presence
   of macrophages in MPS II (a) compared to MPS II (b) and normal controls
   (FigS. [79]1B, S2-B). The recovery of these cells in MPS II (b)
   suggested an improvement in phagocytic function. Additionally, a
   reduction in T cells and B cells in MPS II (b) was observed,
   potentially indicating their depletion following immunoreaction. NK and
   NKT cells remained steady between the control group and two treatment
   stages.

Fig. 1.

   [80]Fig. 1
   [81]Open in a new tab

   Clinical manifestations and single-cell transcriptional profile of
   mucopolysaccharidosis II (MPS-II) patient. (A) Clinical pictures of the
   patient. (a) claw hands of the patient; (b) X-ray image shows the
   j-shape sella; (c) MRI image shows enlarged supratentorial ventricle,
   deepened cerebral sulci, and scattered small cystic with long T1 and
   long T2 signals. (B) Uniform manifold approximation and projection
   (UMAP) plot of integrated data splited by group (Health: normal sample;
   MPS II (a): sample from first stage treatment; MPS II (b): sample from
   second stage treatment), colored by cell type. (C) Cell types
   identified by known marker genes. (D) Heatmap showing top 10
   differential expressed genes in each group. (E) Location of germline
   mutation of IDS, HPSE, HPSE2. (F) Feature plot showing IDS gene
   expression level in three groups. (G) Heatmap illustrating the
   expression level of genes involved in pathways related with IDS in
   three groups.

   DEGs were compared among the control, MPS II (a), and MPS II (b)
   (Figs. [82]1D, S1-C, S1- D, S1-E, Table S-2). Results showed that MPS
   II (a) exhibited higher expression of TNFAIP3, TUBB2A, UBE2S, and
   TRABD2A but lower expression of TYROBP, TRBC1, and TSPO compared to the
   control and MPS II (b) (Fig. [83]1D). TYROBP, TRBC1, and ZNF385A were
   up-regulated in MPS II (b) and the control group, but down-regulated in
   MPS II (a). TYROBP is known to influence immune system dysregulation,
   bone and skeletal abnormalities, and neuroinflammation^[84]36.
   Activation and inactivation of these genes may be associated with ERT.
   GO and KEGG^[85]28 enrichment analyses of DEGs among the healthy
   control, MPS II (a), and MPS II (b) groups were conducted using
   clusterProfiler^[86]26 (Figure [87]S1D, S1-E, Table S-3, S-4). ARSK was
   up-regulated in MPS II (a) but down-regulated in the control and MPS II
   (b). Genes up-regulated in MPS II (a) were enriched in immune-related
   pathways, such as immune cell differentiation and proliferation, Wnt
   signaling pathway, and JAK-STAT signaling pathway (Fig. [88]S1D). In
   contrast, genes up-regulated in MPS II (b) were predominantly enriched
   in inflammatory response, phagocytosis, and immune cell activation
   (Fig. [89]S1E).

   Based on WGS data of the MPS-II PBMCs, several MPS-II-related germline
   mutations were detected, including IDS, HPSE, and HPSE2 (Fig. [90]1E,
   Table S-[91]1). IDS is a key gene associated with MPS-II, involved in
   calcium binding, protein binding, sulfuric ester hydrolase activity,
   hydrolase activity, and heparan sulfate metabolism^[92]37. The
   expression of IDS was high in the early stage (MPS II (a)) but
   decreased in the later stage (MPS II (b)), suggesting that ERT
   suppresses its expression (Fig. [93]1F). Similarly, HPSE and HPSE2,
   which are also involved in GAG metabolism and heparan sulfate
   biosynthesis, show altered expression in MPS-II according to GeneCards
   ([94]www.genecards.org). The expression levels of IDS-related genes
   among the three groups showed that S100A6, CLPX, and GUSB were
   up-regulated in the control and MPS II (b) groups but down-regulated in
   MPS II (a) (Fig. [95]1G).

   Our findings suggest that monocytes and macrophages could play crucial
   roles in the pathogenesis of our MPS-II patient and response to ERT.
   Furthermore, genomic results indicated that IDS mutations accompanied
   by HPSE and HPSE2 were identified in our MPS-II patient.

Monocytes play important roles in response to ERT in MPS-II

   Monocytes, which nearly disappeared in the severe stage (MPS II (a))
   but demonstrated recovery following ERT, were re-clustered into four
   subclusters (Figs. [96]2A, S2-A). Most CD14 + CD16 + monocytes,
   typically classified in the intermediate subset, fell within classical
   (mono1) and nonclassical (mono2) monocyte subclusters, with some cells
   forming a rare subcluster (mono3) associated with the cell cycle,
   differentiation, and trafficking, as well as a subcluster (mono4)
   expressing cytotoxic genes involved in NK and T cell activation^[97]38.
   The average expression of typical marker genes for monocyte subtypes is
   illustrated in Fig. [98]2B. In MPS II (b), monocytes showed
   up-regulated expression of SYNE2, TMCC3, VAV3, TBC1D9, and SSBP2 but
   down-regulated expression of THAP2, TNF, TGFB1, and SEC61G (Figure
   [99]S2E; Table S5). Mono1 (including mono1_1 and mono1_2), which
   mediates proinflammatory reactions such as leukocyte recruitment,
   cytokine and chemokine production, and leukocyte adhesion and
   migration, was decreased in MPS II (b) (Figure [100]S2C). Notably,
   mono1_2 was primarily found in the control samples and was absent in
   MPS II (b) (Fig. [101]2A). The mono2 subcluster, which mediates antigen
   clearance through antigen-antibody complex binding, the mono3
   subcluster, which plays a critical role in the proliferation,
   differentiation, and survivorship of neutrophilic granulocytes, and the
   mono4 subcluster, which eliminates intracellular pathogens, were all
   increased in MPS II (b) (Figure [102]S2C). These dynamic changes in
   monocyte subtypes could be related with disease development.

Fig. 2.

   [103]Fig. 2
   [104]Open in a new tab

   Single-cell transcriptional profile of monocytes. (A) Umap plot of
   monocyte subtypes split by group. (B) Dotplot showing marker genes of
   monocyte subtypes (mono1: classical monocyte; mono2: nonclassical
   monocyte; mono3: Mono3; mono4: Mono4). (C) KEGG^[105]28 enriched terms
   of genes upregulated mono3. (D) KEGG^[106]28 enriched terms of genes
   upregulated mono1_2. (E–F) Visualization of single-cell clusters of
   monocyte using TooManyCells, which was split by samples and cell
   subtypes. Cells begin at the central node and are recursively divided
   based on transcriptional differences. Branch width corresponds to cell
   number. (G) Ridge plot of the distribution of monocyte subtype along
   pesudotime(top), heatmap of differential expressed genes along
   pesudotime trajectory(bottom). Umap plot at right showing the
   trajectory of monocyte subtypes analyzed by monocle.

   Next, pathway enrichment analysis was conducted for mono3 and mono1_2
   due to their specific distribution between the healthy group and MPS II
   (b). Genes up-regulated in mono3 were mainly enriched in neuron-related
   disease pathways, fat-related pathways, and immune-related pathways
   (Fig. [107]2C; Table S6). Genes up-regulated in mono1_2 were enriched
   in neuro-related pathways such as neuroactive ligand-receptor
   interaction and neurodegeneration pathways, as well as immune-related
   pathways such as the IL-17 signaling pathway and cytokine-cytokine
   receptor interaction (Fig. [108]2D, Table S7), which are known to
   function in MPS-II^[109]6. Genes specifically expressed in MPS II (b)
   were involved in fat metabolism, neural diseases and pathways, and
   immune-related pathways (Figure [110]S2D, Table S8), potentially
   elucidating the amelioration of neuropsychiatric symptoms in the
   patient. Other studies have demonstrated the elevation of lipid
   peroxidation and DNA damage in MPS-II patients and the reduction of
   lipid peroxidation after ERT^[111]39, consistent with our data on fat
   metabolism pathways. The monocyte subclusters analyzed using
   TooManyCells^[112]40 clearly classified MPS II (b) as a separate clade,
   with mono3 unique to MPS II (b) and exhibiting aneuploid features
   (Fig. [113]2A and -E, S2-B). Pseudotime^[114]25 and RNA
   velocity^[115]26 analysis revealed a trajectory derived from mono1 and
   mono2 to mono3 and mono4 (Figs. [116]2F and [117]3-D). Dynamic gene
   expression along the monocyte pseudotime trajectory revealed distinct
   patterns. TCF7L2, GNAQ, and SLC8A1 were highly expressed at the
   starting and middle points of the trajectory, corresponding to the
   mono1 subcluster (Fig. [118]2G). Conversely, HES4, WARS, and RHOC were
   predominantly expressed at the beginning of the trajectory, while ADM,
   NLRP3, and CCL20 were highly expressed at the endpoint, associated with
   the mono3 and mono4 subclusters (Fig. [119]2G).

Fig. 3.

   [120]Fig. 3
   [121]Open in a new tab

   Single-cell transcriptional profile of macrophages. (A) Umap plot of
   cell subtypes of macrophages split by group. (B) KEGG^[122]28 terms of
   macrophage in each group. Gene ratio was indicated by the circle size
   and color indicated the adjusted pvalue. (C) Ridgeplot of the
   distribution of macrophage subtypes along pesudotime(top), heatmap of
   differential expressed genes along pesudotime trajectory(bottom). (D)
   RNA velocity analysis delineated dynamic changes in cell fate,
   projected onto a UMAP plot. The right illustrates re-clustered result
   of monocyte and macrophages, the left is RNA velocity plot of monocyte.
   Arrowheads indicated the predicted direction of cell development, with
   arrow size denoting the strength of predicted directionality. (E)
   Visualization of monocyte and macrophage subtypes trajectory using
   TooManyCells. (F) Dot plot of the predicted interactions between
   macrophage subtypes.

   CellPhoneDB^[123]24 was employed to explore cellular interactions among
   the monocyte subtypes. Results revealed that interactions between
   mono1_1 and mono2 were stronger in MPS II (b) compared to the control
   (Figure [124]S2I). Cellular interactions between monocyte subsets
   primarily involved CD74, TNFRSF1B, and C5AR1 (Figure [125]S2H). Mono1_2
   mainly interacted with other monocyte subtypes through CD74 and its
   corresponding protein receptors, such as APP, COPA, and MIF (Figure
   [126]S2G). CD74 encodes a protein associated with class II major
   histocompatibility complex (MHC-II) and serves as a cell surface
   receptor for the cytokine macrophage migration inhibitory factor (MIF),
   which activates T and B cells when bound to the CD74 protein. This
   protein also suppresses the production of amyloid beta when interacting
   with amyloid precursor protein (APP)^[127]41. Results also showed that
   mono2 and mono3 interacted strongly via C5AR1-RPS19 pairs, where RPS19
   acts as an immunosuppressive factor inducing the production of
   immunosuppressive cytokines, including transforming growth factor-β
   (TGF-β) in tumors^[128]42. Reducing RPS19 or blocking the C5AR1-RPS19
   interaction could decrease RPS19-mediated immunosuppression^[129]42.
   Collectively, these results suggest that monocytes may restore immune
   function during ERT of MPS-II patients. Mono3, unique in MPS II (b) and
   found at the terminus of the cellular trajectory, may act as a pivotal
   therapeutic component functioning through the related pathways we
   discovered. Furthermore, the deficiency of mono1_2 may be a significant
   factor in MPS-II, as it is challenging to reactivate through
   therapeutic intervention.

   Overall, our results revealed restoration of monocytes in the recovery
   stage of a MPS II patient (MPS II (b)). Those monocytes were involved
   in inflammation, immune dysregulation, and tissue remodeling and repair
   process. Those results could provide potential therapeutic targets for
   this disease.

Macrophages play an important role in MPS-II treatment

   Four subclusters of macrophages were identified via re-clustering
   analysis. Similar to monocytes, macrophages were almost absent in MPS
   II (a) (Figs. [130]3A, S3-A). Macrophages can be divided into two
   subgroups, M1 and M2, based on their localization and
   environment^[131]43. M1 macrophages are induced by lipopolysaccharide
   (LPS) and interferon (IFN)-γ and secrete proinflammatory cytokines,
   while M2 macrophages are involved in tissue repair and immune
   tolerance, with both able to transform between each other^[132]44. Our
   results showed that YWHAE, STX7, MYADM, and PLAC8 were up-regulated in
   the control group and MPS II (b) but down-regulated in MPS II (a).
   Conversely, PLEK, RNF, NINJ1, and PCNT were up-regulated in MPS II (a)
   but down-regulated in the control group and MPS II (b) (Figure
   [133]S3D, Table S-9). Dynamic expression of these genes indicated their
   response to ERT. Pathway analysis of macrophages indicated that MPS II
   (a) was involved in fatty acid metabolism and biosynthesis and the PPAE
   and NF-κB signaling pathways, while MPS II (b) and the control group
   were enriched in the PI3K-Akt signaling pathway, which was lacking in
   MPS II (a) (Fig. [134]3B, Table S-10). Trajectory analysis of M1 and M2
   macrophages revealed that CITED2, IL32, and MAL were highly expressed
   in M2, while CASC15, MYL9, and THBS1 were up-regulated in M1
   (Fig. [135]3C and -D, S3-C). Among macrophage subtypes, interactions of
   M2-M2 and M2-M1_2 increased significantly in MPS II (b) compared to MPS
   II (a) and the healthy controls, while interactions between M2 and M1_2
   were weaker in MPS II (a) than in MPS II (b) and the healthy controls
   (Figure [136]S3E). These cellular interactions of macrophage subtypes
   revealed that M1_2 and the M2 were strongly connected through PF4-CXCR3
   pairs, a connection not observed in the control or MPS II (a)
   (Fig. [137]3F). CD74 and its corresponding receptors, including MIF,
   COPA, and APP, were strongly associated with M2 and other macrophage
   subtypes in the control and MPS II (b) groups, while these associations
   were absent in MPS II (a) (Fig. [138]3F). Overall, these findings
   suggest that macrophages may play a role in the pathology of MPS-II,
   which could be considered as a potential therapeutic target to mitigate
   disease progression and improve patient outcomes.

   Both monocytes and macrophages exhibited phagocytic function, with
   similar distributions in the MPS-II patient. To investigate the
   relationship between macrophages and monocytes, joint trajectory
   analysis was performed. RNA velocity analysis of the two cells
   indicated potential trans-differentiation from mono1 to mono3
   (Fig. [139]3D). TooManyCells analysis^[140]40 also showed that the
   macrophages were significantly distinct from the monocytes, while mono3
   and mono2 were separated from the distinct mono1 branch (Fig. [141]3E).

   Collectively, the deficiencies of macrophages and monocytes appear to
   be a crucial factor in MPS-II. These two cell types may function
   through biological pathways deficient in MPS II (a) but restored in MPS
   II (b) and the control group. Additionally, ligand-receptor pairs
   interacting between cells may play a potential role in aiding the
   recovery of MPS-II.

Characterization of T cell diversity and clonal expansion in MPS-II

   A total of 39 454 T cells were obtained from the scRNA-seq data,
   annotated into 14 subtypes and one cluster of NKT cells based on
   canonical markers (Figs. [142]4A and B, S4-A, S4-B). The identified
   subtypes included CD4 + Th22, CD4 + memory T cells (CD4 Tm),
   CD4 + regulatory T cells (CD4 Treg), four clusters of CD4 + naïve T
   cells (CD4 Tn), CD8 + Tc17, CD8 + central memory T cells (CD8 Tcm),
   three clusters of CD8 + effector memory T cells (CD8 Tem), and two
   clusters of CD8 + cells (CD8 Tn). After ERT, the proportions of two
   naïve T cell subclusters (CD4 Tn_1, CD8 Tn_1) rebounded. To further
   explore expression changes in these two naïve T cell subtypes, we
   identified DEGs in these cells among the three groups (Figure S4-C,
   S4-D). In the CD4 Tn_1 subtype, TUBB, UCP2, WARS, and TYROBP were
   down-regulated in MPS II (a) but up-regulated in MPS II (b) and the
   control, while ZFP36, TUBB4B, YWHAE, and YES1 were up-regulated in MPS
   II (a) but down-regulated in MPS II (b) and the control (Figure S4-C).
   These genes may play a role in disease development. In the CD8 Tn_1
   subtype, TUBA1B, ZNF331, and TRBV were up-regulated in MPS II (a) but
   down-regulated in MPS II (b) and the control group, while ZYX, VDAC1,
   and TRBV5 were down-regulated in MPS II (a) but up-regulated in MPS II
   (b) and the control group (Figure S4-D). The top 15 DEGs among the
   three groups are shown in Figure S4-E (Table S11). Pathway enrichment
   analysis of T cells revealed that MPS II (a) was enriched in
   neurological disorders, while MPS II (b) was enriched with
   neuroactive-related pathways (Fig. [143]4C, Table S12). These results
   may help explain the improvements in clinical neurological
   manifestations observed in the patient.

Fig. 4.

   [144]Fig. 4
   [145]Open in a new tab

   Single-cell transcriptional profile of T cells and NK cells. (A) UMAP
   plot of cell subtypes of T cell. (B) Barplot of the percentageof T cell
   subtypes in each sample. (C) Bubble plot showing gene ontology (GO)
   terms of each group. Gene ratio was indicated by the circle size and
   color indicated the adjusted pvalue. (D–E) Ridgeplot shows the
   distribution of subtypes of CD4 + T and CD8 + T cell along
   pesudotime(top), heatmap of differential expressed genes along
   pesudotime trajectory(bottom). (F) RNA velocity plot of T cell. (G)
   Umap plot of the distribution of TCR clonotype split by group. (H)
   Alluvial plot illustrating amino acid sequence VDJC genes clonotype
   changes of TCR between samples.

   To further delineate the transition states associated with T cells, the
   trajectories of the CD4 + and CD8 + T cell subtypes were analyzed using
   Monocle2^[146]25 (Figure S5-G, S5-F). CD4 + and CD8 + naïve T cells
   were positioned at the starting point of the trajectory path, whereas
   CD4 Th22, CD4 memory T cells, and two clusters of CD8_Tem were at the
   terminus. The transcriptional changes related to the transitional
   states of CD4 and CD8 T cells were then investigated. The CD4 + and
   CD8 + T cells were categorized into three phases (Fig. [147]4D and -E).
   In CD4 + T cells, phase 1 was characterized by two clusters of CD4
   naïve T cells with up-regulated ITM2A, CHI3L2, SOX4, and LRRN3
   expression. Phase 2 was associated with the up-regulation of KLRF1,
   FGR, TYPOBP, and FCRL3, predominantly featuring CD4 Tm. Phase 3
   consisted mostly of CD4 Th22 cells with up-regulated FAM45A, MT2A,
   TNFRSF4, and FOXP3 expression (Fig. [148]4D). For CD8 + T cells, phase
   1 showed up-regulation of BCL3, NR4A1, and PLK2, dominated by CD8 naïve
   T cells. The second trajectory phase exhibited high expression of MYC,
   SOX4, AIF1, and AQP3 (Fig. [149]4E). RNA velocity analysis indicated
   directional flows from CD4 Treg to CD4 Tm cells, and from CD8 Tcm to
   CD8 Tem cells (Fig. [150]4F). Cellular communication analysis of
   different T cell subtypes revealed enhanced interactions between CD4
   Tn_3 and CD4 Th22 at the MPS II (a) stage, which were weaker in the
   control group and MPS II (b). Conversely, interactions between CD8 Tm_1
   and CD8 Tc17 increased in MPS II (b) compared to the healthy control
   but decreased in MPS II (a) (Figure S4-H, S4-I).

   To determine whether the T-cell repertoire is heterogeneous between the
   two treatment stages, scRNA-seq was combined with TCR-seq data,
   yielding 30 358 T cells (Fig. [151]4G). Expanded T cell clones were
   predominantly found in the CD8 Tem_3 and Tem_1 subclusters (Figure
   S4-J). The top 14 genes in the MPS II (a) clonotypes comprised most of
   those in MPS II (b) (Fig. [152]4H). Notably, 75% of CD8 Tem_3 cells
   were substantially expanded (containing large clonotypes), indicating
   high clonality (Figure S4-J). T cell clonal expansion mainly occurred
   in CD8 Tem cells, with significant expansion of three clonotypes after
   ERT, including RAV1-2.TRAC_TRBV20-1.None.TRBJ2-2.TRBC2T,
   TRAV14/DV4.TRAJ34.TRAC_TRBV19.None.TRBJ1-1.TRBC1, and TRAV29-/DV5.
   TRAJ17.TRAC_TRBV2.None.TRBJ1-1-2.TRBC1 (Fig. [153]4G, H, S4-J). Clonal
   diversity of TCR-seq data demonstrated that the two treatment stages of
   the MPS-II patient were significantly distinct (Figure S4-K). Overall,
   these findings suggest that T cells play a vital role in antigen
   recognition, T cell activation, and adaptive immune response in our MPS
   II patient.

Characterization of B cell diversity in MPS-II

   Re-clustering of the B cell compartment identified six subclusters with
   recognized cell markers, with the B1/2/5 subclusters showing marked
   declines in MPS-II (Fig. [154]5A, S5-A, S5-B, S5-C). Three naïve B cell
   subtypes were clustered together, with one subtype enriched at the MPS
   II (a) stage and decreased in the control and MPS II (b) stages,
   showing up-regulation of PIK3AP1, LYST, and GNG5 (Figure S5-D). Further
   pathway enrichment analysis indicated that MPS-II was enriched in
   ubiquitin-mediated proteolysis, the PI3K-Akt signaling pathway, and
   neurodegeneration-related pathways, while MPS II (a) was enriched in
   nervous system disease pathways (Fig. [155]5B, Table S13). Trajectory
   analysis showed that naïve B-B1 cells were positioned at the beginning
   of the trajectory path with up-regulation of GLIPR1, FAM111B, and GBP4,
   whereas naïve B-B2/3 cells were located at the terminus with high
   expression of LAIR1, SH3BP2, FCRLA, and PLD4 (Fig. [156]5C, S5-F).
   Pseudotime analysis indicated a flow from naïve B cells to memory B
   cells and plasmablasts via scVelo^[157]24 (Fig. [158]5D).
   Ligand-receptor interaction analysis revealed that classical memory B
   cells (B4) strongly interacted with naïve B cell cluster two (B2) in
   the control group and MPS II (b) but showed weaker interactions in MPS
   II (a) (Figure S5-E). In MPS II (b), classical memory B cells displayed
   stronger interactions with naïve B cell cluster three (B3) than in the
   other two groups, as well as more frequent interactions between naïve B
   cell subclusters. Further exploration of ligand-receptor pairs engaged
   in interactions between B cell subtypes (Fig. [159]5E) showed that
   cellular interactions were primarily through CD74 and its corresponding
   proteins (Fig. [160]5E).

Fig. 5.

   [161]Fig. 5
   [162]Open in a new tab

   Single-cell transcriptional profile of B cells. (A) Umap plot of cell
   subtypes of B cell split by groups. (B) Bubble plot showing
   KEGG^[163]28 terms of each group. Gene ratio was indicated by the
   circle size and color indicated the adjusted pvalue. (C) Ridgeplot of
   the distribution of B cell subtype (top) and heatmap of differential
   expressed genes (bottom) along pesudotime trajectory. (D) RNA velocity
   plot of B cell. (E) Dot plot of the predicted interactions between B
   cell subtypes in different group. Interactions are indicated by the
   color. (F) Umap plot illustrating the distribution of BCR clonotype.
   (G) Umap plot of the distribution of BCR clonotype split by group. (H)
   Alluvial plot illustrating amino acid sequence VDJC genes clonotype
   changes of BCR between samples.

   To evaluate the heterogeneity of the B cell repertoire between the two
   treatment stages, scRNA-seq was combined with BCR-seq data, yielding 6
   972 B cells. Expanded B cell clones were predominantly found in naïve B
   cells and B cells from MPS II (b) (Fig. [164]5G and H). Both MPS-II
   stages exhibited clonal expansion, with MPS II (b) showing less intense
   expansion at the end of the treatment stage (Fig. [165]5F and G). Only
   two clonotypes were found in both treatment stages, which decreased at
   the end of the treatment stage, while each stage contained various
   unique clonotypes (Fig. [166]5H). The B cell clonotype
   IGH.185.IGHV1-2IGLC: LD.2.IGLV2-23 shrank following ERT, while
   NA_NA_IGLC: LD.1.IGKV3-15 was only present in MPS II (a). Diversity
   analysis showed that B cells in MPS II (b) exhibited a higher diversity
   index score compared to the other groups (Figure S4-K). These
   clonotypes likely play a crucial role in the ability of B-cells to
   recognize and respond to specific antigens, thereby contributing to the
   adaptive immune response.

ERT influences metabolism and activates the ligand-receptor component for
cell-cell communication in MPS-II

   Cell-cell interactions mediated by ligand-receptor pairs expressed in
   different cell types in the three groups were estimated using
   CellPhoneDB (Fig. [167]6A). The number of interactions between
   different cell types was similar between the control group and MPS II
   (b), both of which were stronger than those in MPS II (a). Notably, the
   interactions between NK, NKT, and CD14 + monocytes increased
   significantly in MPS II (b) compared to the control group. In contrast,
   the cellular interactions between macrophages and other PBMCs were not
   particularly close and remained unchanged across the three groups.

Fig. 6.

   [168]Fig. 6
   [169]Open in a new tab

   Cellular interactions between PBMC cell types. (A) Number of cell-cell
   interactions between each cell types in each group. (B) KEGG pathway
   dotplot of health control, MPS1, and MPS2. (C–E). Dot plot of the
   predicted interactions between each cell types in different group.
   Interactions are indicated by the color. (C) ligan-receptor pairs
   between cells except monocyte and macrophage; (D) ligan-receptor pairs
   between monocyte and other cell types; (E) ligan-receptor pairs between
   macrophage and other cell types. (F) Predicated cell transformation and
   cellular interactions through potential ligand-receptor pairs.

   Monocytes and macrophages were specifically scarce in MPS II (a),
   indicating a potential mechanism involving interactions between these
   cells and others that impact therapeutic responses during ERT in
   MPS-II. To investigate this, we first explored the differences in
   enriched pathways between different treatment stages. Notably, the Wnt,
   prolactin, and adipocytokine signaling pathways were enriched in MPS II
   (a) but absent in the other stages (Fig. [170]6B). Genes related to the
   TNF and JAK-STAT signaling pathways were more prevalent in MPS II (a)
   compared to MPS II (b) (Fig. [171]6B), consistent with the increased
   plasma and serum levels of TNF-α cytokines in MPS-II patients^[172]45,
   demonstrating the effectiveness of ERT. Dysregulation of the JAK-STAT
   signaling pathway is associated with various cancers and autoimmune
   diseases^[173]46. Additionally, the FoxO signaling pathway, which is
   important in metabolism, cellular proliferation, stress resistance, and
   apoptosis, was enriched in MPS II (a) but absent in the control group
   and MPS II (b) (Fig. [174]6B). Neural-related pathways, such as the
   sphingolipid signaling pathway, serotonergic signaling pathway,
   neuroactive ligand-receptor interaction, neurotrophin signaling
   pathway, glutamatergic synapse, and GABAergic synapse, were enriched in
   MPS II (b) (Fig. [175]6B), consistent with the improved neurological
   symptoms observed in the patient.

   To further explore the cellular interactions between monocytes and
   macrophages with other cells, receptor-ligand interactions between
   different cell types were predicted (Fig. [176]6C, D and E). The
   interaction of NK cells with other immune cells was primarily mediated
   by HLA-E and their corresponding protein receptors, which were enhanced
   in MPS II (b) but absent in MPS II (a) and the control group
   (Fig. [177]6C). CD74 and its corresponding receptors, mediating
   interactions between B cells and T cells, exhibited higher expression
   in MPS II (b) compared to MPS II (a) and the control group.
   Interactions between monocytes and other cell types were mainly
   mediated by ILB1, ANXA1, LGALS9, and their corresponding protein
   receptors, which are associated with anti-inflammatory activity
   (Fig. [178]6D). The IL1B-ADRB2 and C5AR1-RPS19 pairs were significantly
   enriched in monocyte and other cell types. Animal studies have shown
   that activation of ADRB2 can attenuate monocyte activation and
   proinflammatory and pro-fibrotic responses^[179]47. C5AR1 is an
   upstream signal in the Toll-like pathway, which can activate downstream
   inflammatory cytokines^[180]48, such as TNFRA family members (TNFAIP6,
   TNFRSF1B, and TNFSF10) and IL family members (IL1B, IL7R, and
   IL18R1)^[181]34,[182]35. Macrophages were primarily associated with B
   cells through CD74 across all three groups and with NK cells through
   PPBP in MPS II (a) and MPS II (b) (Fig. [183]6E).

   In conclusion, the transformation of mono3/4 from mono1/2 and their
   interactions with NK cells through the IL1B-ADRB2 pair and B cells
   through the TNFSF13B-HLA-DPB1 pair were revealed in our results.
   Although only one subject remains controversial, our results also could
   propose a possible atlas of cellular interaction network during ERT in
   MPS II (Fig. [184]6F).

Discussion

   In our study, we explored the transcriptional and genomic changes of
   MPS-II patients during ERT. Our findings revealed up-regulation of
   TNFAIP3 in MPS II (a), the early stage of ERT, and down-regulation of
   this gene in the end stage of ERT. The protein encoded by TNFAIP3
   inhibits NF-κB activation and TNF-mediated apoptosis^[185]49.
   Accumulation of the GAG heparan sulfate in conditions such as MPS-I and
   MPS-II may trigger the production and release of proinflammatory
   cytokines like TNF-α^[186]50,[187]51. Elevated TNF-α levels were
   overserved in MPS VII and related with severe somatic pain^[188]52.
   Prior research has also demonstrated that anti-TNF-α is effective in
   human studies of MPS I and MPS II, and in mouse models of MPS VI and
   MPS VII^[189]53,[190]54. The transcriptional changes in TNFAIP3 during
   different ERT stages may be associated with our patient’s improvement.
   Genes up-regulated in MPS II (b) and the control group but
   down-regulated in MPS II (a), such as TSPO, TKT, WARS, and TYMP, may
   also play a role in the therapeutic response to ERT.

   MPS-II results from a deficiency of iduronate-2-sulfatase encoded by
   the IDS gene. Our research showed that genes involved in pathways
   related to IDS, such as CLPX, GUSB, and DSC3, exhibited changes at
   different treatment stages, suggesting potential as therapeutic targets
   for MPS. Furthermore, our findings indicated that monocytes and
   macrophages were notably absent or scarce in MPS II (a) (the severe
   stage), likely due to their migration from peripheral blood to tissues
   with accumulated “GAGs” to eliminate these deposits. This suggests that
   these cell types may represent therapeutic targets during ERT. The
   mono3 subcluster was unique to MPS II (b), located at the terminal
   point of the trajectory from mono1 and mono2, suggesting that the
   transition of monocytes is closely associated with disease recovery.
   Genes up-regulated in mono3 were primarily enriched in neuron-related
   pathways, including neurodegeneration and Alzheimer’s disease,
   immune-related pathways, including T cell differentiation and antigen
   processing, and fatty-related pathways, which may be related to the
   improvements observed in the neurological and immune systems of the
   patient. CD4 naïve T cells were reduced in MPS II (a) and MPS II (b)
   compared to the healthy controls, possibly due to the activation of
   naïve T cells by metabolic dysfunction of “GAGs” and infection in the
   patient, or developmental disability of thymus T cells caused by the
   disease itself.

   Targeted treatment of inflammation to attenuate neurodegeneration may
   represent a novel therapeutic direction for patients undergoing ERT,
   which has limited effects on the nervous system. One subtype of mono1
   (mono1_2) was scarce in MPS II (b), showing enrichment in genes
   involved in the Toll-like receptor signaling pathway, TNF signaling
   pathway, and IL-17 signaling pathway. T cells in MPS II (b) also showed
   enrichment in genes involved in the Toll-like receptor and TNF
   signaling pathways. Microtubulation of phagosomes and lysosomes induced
   by Toll-like receptor 4 (TLR4) signal transduction plays an important
   role in antigen presentation^[191]50. The accumulation of GAGs can
   overactivate TLR4, leading to the up-regulation of TNF-α and IL-1β,
   which are significantly down-regulated upon TLR4 inhibition, indicating
   that the TLR4-mediated inflammatory response may play a role in
   MPS-IIIA disease^[192]52,[193]55,[194]56. Elevated TNF-α levels are
   associated with severe somatic pain and impaired physical function in
   patients with MPS-I, -II, and -VI^[195]57. IL-1 can enhance
   inflammatory responses and neurodegenerative diseases^[196]58. GAGs
   bind to TLR4 receptors and activate the TLR4 pathway, leading to the
   release of cytokines such as TNF-α and IL-1β. Inactivation of the TLR4
   pathway in MPS-VII mice lacking TLR4 alleviates clinical and
   pathological features, while anti-TNF-α drug administration in MPS-VI
   rats reduces inflammatory cytokine levels and improves joint
   pathology^[197]59. Furthermore, treatment with pentosan polysulfate
   (PPS), an anti-TNF-α medication, can improve joint motion range in
   adult patients with MPS-II^[198]60. In conclusion, our findings suggest
   that the TLR4/TNF-α signaling pathway may be implicated in MPS-related
   osteoarthropathy, and the TLR4/IL-1 signaling pathway may play a role
   in neurodegeneration.

   Interactions between monocytes and other cell types increased in MPS II
   (b), particularly with NK and NKT cells, primarily mediated through
   IL1B-ADRB2 pairs. Interactions between B cells and monocytes were
   predominantly mediated by HLA-DPB1-TNFSF13B pairs, which also increased
   in MPS II (b), while monocytes and T cells connected through
   C5AR1-RPS19 pairs. Interactions between monocytes and macrophages were
   closer than with other cells, mainly mediated through ADORA2A-NAMPT
   pairs that only existed in MPS II (b). The mono3 subtype of monocytes,
   which was unique to MPS II (b), could transform from other monocyte
   subtypes (mono1 and mono2). The connection with mono2 through
   C5AR1-RPS19 pairs may play a role in the function of ERT, contributing
   to prognosis.

   However, our studies have certain limitations, included small sample
   size, lacking of long-term observation, and incomplete clinical data of
   biomarkers (GAGs, pro-inflammatory factors, etc.) from different
   specimens (urine, blood, and CSF). Considering the broad heterogeneity
   within MPS, more samples of different types of MPS should be brought
   into our future studies.

Conclusion

   Our study presents a comprehensive transcriptional and genomic atlas of
   MPS-II, utilizing multi-omics approaches and scRNA-seq to investigate
   dynamic changes during ERT treatment. Our findings highlight the
   crucial roles of monocytes and macrophages in ERT, with their
   restoration at the end stage suggesting involvement in pathways such as
   NOD-like receptor and PPAR signaling, and interactions with NK cells, T
   cells, and B cells to improve prognosis. However, the lack of patient
   samples and animal models limits further exploration of the detailed
   molecular mechanisms underlying MPS-II pathology. Future research
   should focus on enrolling more patients and developing animal models to
   enhance our understanding of MPS-II.

Electronic supplementary material

   Below is the link to the electronic supplementary material.
   [199]Supplementary Material 1^ (1.6MB, pdf)
   [200]Supplementary Material 2^ (60.3KB, pdf)
   [201]Supplementary Material 3^ (112.6KB, xls)

Acknowledgements