Relative acidic compartment volume as a lysosomal storage disorder-associated biomarker

Abstract

Lysosomal storage disorders (LSDs) occur at a frequency of 1 in every 5,000 live births and are a common cause of pediatric neurodegenerative disease. The relatively small number of patients with LSDs and lack of validated biomarkers are substantial challenges for clinical trial design. Here, we evaluated the use of a commercially available fluorescent probe, Lysotracker, that can be used to measure the relative acidic compartment volume of circulating B cells as a potentially universal biomarker for LSDs. We validated this metric in a mouse model of the LSD Niemann-Pick type C1 disease (NPC1) and in a prospective 5-year international study of NPC patients. Pediatric NPC subjects had elevated acidic compartment volume that correlated with age-adjusted clinical severity and was reduced in response to therapy with miglustat, a European Medicines Agency–approved drug that has been shown to reduce NPC1-associated neuropathology. Measurement of relative acidic compartment volume was also useful for monitoring therapeutic responses of an NPC2 patient after bone marrow transplantation. Furthermore, this metric identified a potential adverse event in NPC1 patients receiving i.v. cyclodextrin therapy. Our data indicate that relative acidic compartment volume may be a useful biomarker to aid diagnosis, clinical monitoring, and evaluation of therapeutic responses in patients with lysosomal disorders.

Figures

Figure 1. Relative LE/Lys volume increases progressively in B cells in a mouse model of NPC1 disease and parallels the rate of GSL storage in the brain.

(A and B) Splenic (A) and circulating (B) B cells were analyzed by flow cytometry over the lifespan of the Npc1–/– mouse (3 weeks, presymptomatic; 6 weeks, early symptomatic; 9 weeks, late symptomatic) and compared with wild-type Npc1+/+ controls. A pan–B cell marker was detected using an anti-CD19 monoclonal antibody conjugated to PE, and cells were costained with Lysotracker green. Analysis was performed on gated B cells (CD19+), and Lysotracker staining levels were standardized to fluorescent microbeads conjugated to defined numbers of fluorescent molecules/bead to linearize the logarithmic data collected by the cytometer. There was progressive elevation in B cell MEFL at each time point relative to the previous age point (P < 0.01, spleen, all ages, and blood, 6 and 9 weeks). (C and D) GSL analysis, performed by HPLC, of GM2 ganglioside (C) and lactosylceramide (D) in the forebrain of 2-, 4-, 7-, and 9-week-old Npc1–/– mice compared with wild-type controls. **P < 0.01; ***P < 0.001.

Figure 2. Relative LE/Lys volume identifies pediatric patients with NPC1 disease.

Data were collected prospectively on NPC1 patients (n = 97, aged 2 months to 53 years), heterozygous carriers (NPC Het; n = 40, aged 23 to 76 years), age-matched controls (n = 53, aged 4 months to 64 years), and 2 pediatric patients with Tay-Sachs disease (TS). (A) Lysotracker values from circulating B cells were measured by flow cytometry and plotted against patient age. Each patient only appears once, standardized to the first sample analyzed for each patient. ***P < 0.01. (B) When the same patients’ data were plotted as unadjusted severity score (with higher value denoting greater disease burden) against age, multiple subgroups were identified. These were designated according to decreasing severity, from the severest (subgroup a) to the mildest (subgroup f). Rate of change per year is indicated for each subgroup. (C) Total severity score (without hearing, as this was not measured by all clinical centers) for each patient was divided by patient age to generate ASIS. When plotted against MEFL, a statistically significant correlation was found (P = 0.017). (D) When each patient’s plasma cholestane-3β,5α,6β-triol level was plotted against their MEFL value, a statistically significant correlation was observed in all patients (P < 0.01). A piecewise linear model significantly improved the goodness-of-fit with respect to a single linear model (P < 0.001); in this model, the initial slope was significant (P = 0.015), and the second was not (P = 0.228), with the estimated change point being 21,269 MEFL.

Figure 3. Miglustat treatment reduces lysosomal storage in NPC1 patients’ B cells.

Box plots summarizing progression of storage for untreated (n = 7), chronic miglustat (n = 17), and pre-post miglustat (n = 14) patient groups as well as adult controls (n = 14) analyzed over time. The box is limited by the 1st and 3rd quartiles (Q1 and Q3), and the median is indicated by a bar inside the box; whiskers and outliers are defined in terms of 2 unmarked boundaries defined using Tukey’s rule as Q1 – 1.5 IQR and Q3 + 1.5 IQR (IQR, interquartile range, i.e., Q3 – Q1). Whiskers extend to the smallest and largest observed values within the boundaries, and outliers are defined as points outside the boundaries. Adjusted pairwise multiple comparisons are shown (Wilcoxon-Mann-Whitney test). See Table 1 and Supplemental Figure 7.

Figure 4. Relative LE/Lys volume is a sensitive measure of response and potentially adverse response to NPC disease therapeutics.

(A) Pre-post analysis of an NPC2 patient with BMT, using Lysotracker analysis of circulating B cells. MEFL was reduced 1 month after BMT, and by 3 months was within the pediatric control range. At 3 months after BMT, donor chimerism developed and was detected by the presence of 2 B cell populations (inset; gated on mononuclear cells). The populations had high or low Lysotracker fluorescence, with a 50:50 distribution of B cells between the 2 populations. HLA typing confirmed the development of 50% donor chimerism (not shown). (B) 7 patients on HPβCD therapy (individual use INDs) were analyzed. Patients 1–4 were analyzed pre-post i.v. HPβCD (arrows denote therapy initiation points), patients 5 and 6 were analyzed after HPβCD only, and patient 7 was analyzed pre-post intrathecal HPβCD delivery (no i.v. administration throughout the treatment period). (C) B cells isolated from the blood of patients 1–4 were analyzed to determine their levels of GSL storage. Total GSL levels were compared before and after HPβCD delivery for each patient and plotted individually (blue; error bars derived from 3 independent HPLC analytical runs of the same samples). An average of the 4 patients was also plotted (green; error bars derived from the mean values for each patient). *P < 0.05; ***P < 0.001.

Figure 5. Summary of major findings in the study.