Investigating the role and regulation of GPNMB in progranulin-deficient macrophages

Abstract

Introduction: Progranulin (PGRN) is a holoprotein that is internalized and taken to the lysosome where it is processed to individual granulins (GRNs). PGRN is critical for successful aging, and insufficient levels of PGRN are associated with increased risk for developing neurodegenerative diseases like AD, PD, and FTD. A unifying feature among these diseases is dysregulation of peripheral immune cell populations. However, considerable gaps exist in our understanding of the function(s) of PGRN/GRNs in immune cells and their role in regulating central-peripheral neuroimmune crosstalk. One of the most upregulated genes and proteins in humans with GRN haploinsufficiency and in aged Grn knock-out (KO) mice is glycoprotein non-metastatic B (GPNMB) but its normal role within the context of immune crosstalk has not been elucidated.

Methods: To address this gap, peritoneal macrophages (pMacs) from 5-to-6-month old WT and Grn KO mice were assessed for Gpnmb expression and stimulation-dependent cytokine release in the presence or absence of the Gpnmb extracellular domain (ECD). Cellular localization, as well as inhibition of, the microphthalmia-associated transcription factor (MITF) was assessed to determine its mechanistic role in Gpnmb overexpression in Grn KO pMacs.

Results: We observed an increase in GPNMB protein and mRNA as a result of insufficient progranulin in peripheral immune cells at a very early age relative to previous reports on the brain. Stimulation-dependent cytokine release was decreased in the media of Grn KO pMacs relative to WT controls; a phenotype that could be mimicked in WT pMacs with the addition og GPNMB ECD. We also found that MITF is dysregulated in Grn KO pMacs; however, its nuclear translocation and activity are not required to rescue the immune dysregulation of Grn KO macrophages, suggesting redundancy in the system.

Discussion: These findings highlight the fact that knowledge of early-stage disease mechanism(s) in peripheral populations may inform treatment strategies to delay disease progression at an early, prodromal timepoint prior to development of neuroinflammation and CNS pathology.

Keywords: GPNMB; MITF; inflammation; macrophage; progranulin.

Figure 1

GPNMB protein is overproduced in Grn KO pMacs (A) Representative western blot images of GPNMB, PGRN, and GAPDH from B6 and Grn KO pMacs (B) Quantification of PGRN signal normalized to GAPDH. (C) Quantification of GPNMB signal normalized to GAPDH, n=6 per sex and genotype, n = 5-6 per sex and genotype, unpaired t-test, bars represent mean +/- SEM. (D) qPCR data on baseline Gpnmb transcript from bothB6 and Grn KO pMac, n=5-7 per sex and genotype, two-way ANOVA with Sidak’s multiple comparisons test. (E) GPNMB ELISA measurement of GPNMB ECF in culture media generated by pMacs over 48 hours, n=4-5 per sex and genotype, two-way ANOVA with Sidak’s multiple comparisons test. Letter(s) above the bar graphs represent the results of the post-hoc tests. Groups that share the same letter are not significantly different from one another.

Figure 2

Grn KO pMacs display reduced LPS-dependent expression of early inflammatory genes (A) qPCR for Il-1b expression in female B6 and Grn KO pMacs. (B) qPCR for Il-6 expression in female B6 and Grn KO pMacs. (C) qPCR for Tnf expression in female B6 and Grn KO pMacs. (D) qPCR for Il-1b expression in male B6 and Grn KO pMacs. (E) qPCR for Il-6 expression in male B6 and Grn KO pMacs. (F) qPCR for Tnf expression in male B6 and Grn KO pMacs. (G) qPCR for Il-10 expression in female B6 and Grn KO pMacs. (H) qPCR for Il-10 expression in male B6 and Grn KO pMacs. (I, J) GPNMB ELISA measurement of GPNMB ECF in culture media generated by pMacs during experimental paradigm. Two-way ANOVA with Sidak’s multiple comparisons test for every panel, n = 5-7, bars mean represent +/- SEM. Groups that share the same letter are not significantly different from one another.

Figure 3

Re-addition of progranulin protein rescues GPNMB dysregulation in Grn KO pMacs (A) Representative western blot images of GPNMB and GAPDH from female B6 and Grn KO pMacs with and without 48 hours of 4ug/mL of recombinant progranulin treatment. (B) Representative western blot images of GPNMB and GAPDH from male B6 and Grn KO pMacs with and without 48 hours of 4ug/mL of recombinant progranulin treatment. (C) Quantification of GPNMB signal normalized to GAPDH. (D) Quantification of GPNMB signal normalized to GAPDH. Two-way ANOVA with Sidak’s multiple comparisons test, n = 6 per genotype, bars represent mean +/- SEM. Groups that share the same letter are not significantly different from one another.

Figure 4

MITF localization is primarily nuclear in Grn KO pMacs and unresponsive to progranulin re-addition. (A) Representative western blot images of MITF, GAPDH, and H3 from B6 and Grn KO pMacs with and without 48 hours of 4ug/mL of recombinant progranulin treatment. (B) Quantification of MITF signal normalized to GAPDH signal. (C) Quantification of MITF signal normalized to H3 signal. Two-way ANOVA with Sidak’s multiple comparisons test, n = 6 per genotype, bars represent mean +/- SEM. Groups that share the same letter are not significantly different from one another.

Figure 5

Recombinant progranulin is processed with faster kinetics in Grn KO pMacs relative to WT pMacs. (A) Representative western blot images of PGRN and GAPDH signal from pMacs collected from female B6 and Grn KO with and without 48 hours of 4ug/mL of recombinant progranulin treatment. (B) Representative western blot images of PGRN and GAPDH signal from pMacs collected from male B6 and Grn KO with and without 48 hours of 4ug/mL of recombinant progranulin treatment. (C) Quantification of PGRN signal normalized to GAPDH signal from female B6 and Grn KO pMacs with and without 48 hours of 4ug/mL recombinant progranulin treatment. (D) Quantification of PGRN signal normalized to GAPDH signal from male B6 and Grn KO pMacs with and without 48 hours of 4ug/mL recombinant progranulin treatment. Two-way ANOVA with Sidak’s multiple comparison test, n=6 per genotype, per treatment, bars represent mean +/-SEM. Groups that share the same letter are not significantly different from one another.

Figure 6

Inhibition of MITF activity alone does not rescue the overproduction of GPNMB transcript or protein in Grn KO pMacs (A) Representative western blot images of B6 and Grn KO female pMacs treated with vehicle (DMSO), 5uM, or 10uM of ML329 for 24 hours. (B) Representative western blot images of B6 and Grn KO male pMacs treated with vehicle (DMSO), 5uM, or 10uM of ML329 for 24 hours. (C) Quantification of GPNMB signal normalized to GAPDH from female pMacs, n=-5-6 per genotype and treatment. Two-way ANOVA with Sidak’s multiple comparisons test. (D) Quantification of GPNMB signal normalized to GAPDH from male pMacs, n=-5-6 per genotype and treatment, bars represent mean +/- SEM. Two-way ANOVA with Sidak’s multiple comparisons test. (E) qPCR for Gpnmb expression in female B6 and Grn KO pMacs, n=5-8 per genotype per treatment. (F) qPCR for Gpnmb expression in male B6 and Grn KO pMacs, n=5-8 per genotype per treatment, bars represent mean +/-SEM. Two-way ANOVA with Sidak’s multiple comparisons test. (G) qPCR for Ctsd expression in female B6 and Grn KO pMacs, n=5-8 per genotype per treatment, bars represent mean +/-SEM. (H) qPCR for Ctsd expression in male B6 and Grn KO pMacs, n=5-8 per genotype per treatment, bars represent mean +/-SEM. Two-way ANOVA with Sidak’s multiple comparisons test. Groups that share the same letter are not significantly different from one another.