Immunoregulators Multitasking in Neurons is not Just a WHIM

Abstract

Distinguishing direct from indirect mechanisms is often difficult in multisystem genetic conditions. In this issue of Neuron, Demenego et al.¹ report that cerebellar malformation and dysfunction in WHIM syndrome immunodeficiency result from direct effects of hyperfunctional CXCR4 in neurons.

Phylogenetic diversification is accomplished using a relatively uniformly-sized exome. For example, H. sapiens has an adaptive immune system and a brain containing ~85 billion neurons, whereas the nematode C. elegans lacks adaptive immunity and has only 302 neurons; yet, both species are coded by ~20,000 genes.² Phenotypic diversification requires emergence of specialized sets of genes, but may also result from acquisition of additional functions by individual genes. In this regard, Mendelian disorders often present as multisystem syndromes, providing a platform to identify candidate multitasking genes in humans.

In particular, many inborn errors of immunity (IEIs) are associated with neurologic dysfunction; however, distinguishing direct neuronal mechanisms from indirect effects of immune dysfunction is challenging. Demenego et al. have taken on this challenge in the case of WHIM syndrome, an ultrarare IEI caused by gain-of-function mutations in the G protein-coupled chemokine receptor CXCR4, through study of the nervous system in Cxcr4^+/1013 WHIM model mice.^3,4 Most WHIM mutations are in the CXCR4 C-tail and truncate the receptor, removing sites for agonist-induced phosphorylation by G protein-coupled receptor kinases. Normally, the phosphorylated tail of activated CXCR4 docks β-arrestins to mediate receptor internalization and downregulation, terminating signaling. Thus, paradoxically, loss of WHIM receptor structure increases function.

The acronym WHIM conveys the highly penetrant immunologic phenotypes in patients (warts, hypogammaglobulinemia, infections and myelokathexis [neutropenia despite myeloid hyperplasia from retention of neutrophils in bone marrow]), but omits less penetrant extrahematopoietic phenotypes, including developmental abnormalities of the urogenital, cardiovascular and nervous systems. Excepting Tetralogy of Fallot, extrahematopoietic WHIM phenotypes are asymptomatic or subtle. In the nervous system, in particular, 4 of 6 WHIM patients imaged had defective cerebellar foliation, with abnormal orientation of the gracilis and biventer lobules and tonsils, but without overt neurologic symptoms.⁵ Psychomotor symptoms were diagnosed by specific testing in 3 subjects, including ‘fine and global motor coordination disorders, balance disturbances, mild limb ataxia and excessive talkativeness’, and ‘increased risk of internalizing and/or externalizing problems’. The psychiatric findings reached a clinical level for only 5 of 66 measurements. Since the patients ranged in age from 8–51, defining whether the psychomotor findings might be independent of immune dysfunction and the stress of recurrent infection and inflammation was not possible.

CXCR4 is normally constitutively expressed in most types of leukocytes, where it coordinates homeostatic leukocyte trafficking and acts as a retention signal for leukocytes developing in or homing to bone marrow niches, as well as in some non-hematopoietic cell types, including the granule cell lineage in cerebellum.^3,6,7 In this regard, Cxcr4 knockout mice, which are non-viable, have defective embryonic granule cell migration resulting in an abnormal external granule cell layer, as well as misplaced Purkinje cells and axonal disorganization.^6,7 Finding cerebellar malformation in WHIM patients and now prenatally in classic Cxcr4^+/1013 WHIM model mice supports the notion that precise CXCR4 signaling fine-tunes normal cerebellar development by multitasking directly in cerebellum. The viability of WHIM mice allowed a wealth of new insights and questions.

At the molecular level, transcriptome analysis revealed major distortion of gene expression across the pseudotime trajectory of WHIM model mouse granule cell differentiation, but not in microglial cells, highlighting the question of whether granule cell precursor migration and differentiation defects independently result from increased CXCR4 signaling or are instead linked.

At the cellular level, cultured cerebellar granule cells from WHIM mice generated neurons that had increased branching, and increased spike frequency and spike synchronicity rate, suggesting direct granule cell-autonomous effects of hyperfunctional CXCR4 on migration, excitability and synaptic plasticity. However, it remains unclear whether the observed defects result from constitutive hyperfunctional CXCR4 activity versus tonic autocrine or paracrine stimulation of CXCR4 by CXCL12 expressed by the cells versus epigenetic effects on cell state established in vivo prior to explant.

At the anatomic level, the most pronounced malformation was in vermis, which had increased volume along the primary fissure, decreased volume along the preculminary fissure and decreased white matter arbor vitae LIII and LIX branch length. The length of the Purkinje cell layer was decreased in several lobules, including the paramedian lobule, which in rodents corresponds to the human gracilis lobule, the most prominently malformed region in the vermis of WHIM patients.⁵

At the behavioral level, WHIM mice had normal spontaneous locomotion in ambulation and air righting tests but displayed increased thigmotaxis in the open field test, spending 50% less time than wild type littermates in the center and more time close to the walls, while traveling more slowly and less far. This was judged to be increased anxiety-like behavior, not motor dysfunction, since performance on the rotorod test was normal and there was reduced exploratory behavior in the dark-light chamber test and decreased immobility time in the forced swim test. Anxiety-like behavior here aligns with increased recognition that the cerebellum contributes to anxiety.⁸ It also aligns loosely with the psychiatric findings in some WHIM patients, with the caveat that patient phenotypes were not established as more common and/or severe than in the general population.⁵ When challenged using a battery of additional standardized behavioral tests, WHIM pups performed less well than wild type littermates on the open field traversal and surface righting tests, the forelimb grasp test of neuromuscular development, the negative geotaxis test of vestibular function, the ear twitch and auditory startle tests of sensory function and the cliff aversion test of sensory/motor coupling. There are no data available in WHIM patients during comparable early stages of physical, sensory, and motor development.

At the pharmacologic level, remarkably, intraventricular injection of one microgram of the Cxcr4 antagonist AMD3100 (plerixafor) at either day E12.5 or P1 improved both cerebellar defects and behavioral test performance (cliff aversion and open field tests), measured at day P45 and P7. AMD3100 has a half life of only ~5 hours when injected subcutaneously, resulting in briefly increased white blood cell counts in wild type and WHIM mice as well as healthy human subjects and WHIM patients,⁹ so finding a durable effect on neurologic phenotypes in WHIM mice given a single dose is quite exceptional and worth further investigation of time course, dose-response and importance of the intraventricular route of administration, as well as mechanistic study of direct protein expression and function of CXCR4 on granule precursor cells, local expression and regulation of CXCR4’s sole chemokine ligand CXCL12, and potential modulation by cerebellar expression of the atypical CXCL12 receptor ACKR3.

Similarly, the data do not exclude contributions by other brain regions and cell types, including neurons, glia and resident or circulating immune cells, on cerebellar and psychomotor phenotypes, despite no obvious difference in cell number and distribution in brain compared to wild type control mice, since other cell types might also have an altered transcriptome and functional state at critical developmental timepoints that were not assessed, and could be affected by AMD3100 blockade of CXCR4, either directly or indirectly. Refining mechanistic understanding of the neurologic phenotypes in WHIM mice would benefit from analysis of chimeric wild type/WHIM mice generated either by bone marrow transplantation or ideally by engineering mice to express a WHIM allele of CXCR4 in a controllable, granule cell precursor-specific manner. To date, other extrahematopoietic phenotypes observed in WHIM patients have not been reported in WHIM mice.

Immunologists receive two bonuses from this paper. The first is detailed analysis of immune system development in utero and in the perinatal period in WHIM mice, finding similar hematologic defects as those reported in mature WHIM mice and patients.^3,4 The blood findings agree with limited data in neonatal patients and were extended to liver and spleen, where patient data are lacking. The second bonus is comprehensive analysis of neurologic phenotypes and regional brain expression of disease genes for 517 annotated IEIs. Mining existing databases, the authors found that 41.6% of IEIs have neurologic phenotypes and that IEI gene expression directly in brain could be classified in distinct spatio-temporal categories, from embryonic to post-natal/adult, and by predominant brain region and cell type, including neurons. Like CXCR4, cerebellar expression predominated for other IEI genes, with thalamus a distant second. Single cell analysis suggests that many of these immunoregulators may also multitask in neurons providing a roadmap for future work.

The authors conclude, reasonably, that cerebellar malformation and psychomotor defects in WHIM mice arise from hyperfunctional Cxcr4 activity in neurons during development and can be reversed by early single-dose intraventricular treatment with a CXCR4 antagonist. Thinking of WHIM syndrome as not just an immunological disorder with neurologic symptoms but instead as both a primary immunological disorder and primary neurological disorder is conceptually important (Figure 1). The effect sizes found in WHIM mice by behavioral testing suggest that CXCR4 serves mainly to fine tune other more dominant developmental and functional processes. In this regard, strong and overt psychomotor dysfunction has not been reported as a spontaneous phenotype in WHIM patients despite lifetimes of recurrent infection and inflammation.⁵ Thus, the elegant and insightful results from this work should motivate study of additional patients to more broadly and precisely define the penetrance and range of clinical severity and importance of neurologic phenotypes in WHIM syndrome.

Figure 1.

Direct CXCR4 regulation of immunity, cerebellar development and psychomotor function. Gain-of function mutations in CXCR4 cause WHIM syndrome, a primary immunodeficiency disease and primary neurologic disorder manifesting as cerebellar malformation. WBC, white blood cell. The figure was created using Biorender.