Olfaction in Homozygous and Heterozygous SYNJ1 Arg258Gln Mutation Carriers

View the entire article with references and supplemental information on the Wiley Online Library. 
Note:  Reference links embedded in the article below will also take you to the article on the Wiley Online Library.

Return to Table of Contents

Authors:  Marina Picillo, Anna De Rosa, Maria Teresa Pellecchia, Chiara Criscuolo, Marianna Amboni, Roberto Erro, Vincenzo Bonifati, Giuseppe De Michele and Paolo Barone

Article first published online:   25 JUL 2015 | DOI: 10.1002/mdc3.12183


Hyposmia is a common nonmotor symptom in Parkinson's disease (PD) and has been variably detected in monogenic parkinsonism. SYNJ1 has been recently identified as the gene defective in a novel form of autosomal-recessive, early-onset atypical parkinsonism, designed as PARK20. To assess olfaction in PARK20, we administered the University of Pennsylvania Smell Identification Test (UPSIT) in four groups of subjects: SYNJ1 homozygous (HOM = 3) and heterozygous (HET = 4); sporadic PD (PD = 68); and healthy control subjects (CTR = 61). A linear regression model was constructed to assess the association between raw UPSIT score (outcome) and group (HOM, HET, PD, and CTR), adjusting for age, gender, and current smoking status. Likewise in PD patients, odor identification is impaired in homozygous SYNJ1 mutation carriers. Although the limited sample size precludes definite conclusions about olfaction in SYNJ1-related parkinsonism, our findings suggest new insights into PARK20 phenotype and pathophysiology.

SYNJ1 has been recently identified as the gene defective in a novel form of autosomal-recessive, early-onset atypical parkinsonism.[1]Three families (two of Italian and one of Iranian origin) with the same SYNJ1 homozygous c.773G>A mutation segregating with early-onset atypical parkinsonism have been discovered to date.[1-3] The SYNJ1 gene encodes synaptojanin 1, a phosphoinositide phosphatase protein involved in regulation of postendocytic recycling of synaptic vesicles. This form has been designed as PARK20 (Online Mendelian Inheritance in Man, OMIM 615530). In all the PARK20 patients reported on thus far (n = 6), the first symptoms of parkinsonism developed during the third decade of life. Cognitive decline, history of seizures, and supranuclear gaze palsy have been variably observed in this phenotype. Response to levodopa was generally poor and limited by the development of severe and disabling dyskinesias or postural hypotension.[1-3] The 3-year follow-up of the original Italian siblings suggested that SYNJ1-related early-onset atypical parkinsonism may present a more severe progression in the early stages and then reach a relative stability in the later stages.[4]

Impairment in olfactory function is considered a sign of neurodegeneration,[5] and, as such, it is one of the most common nonmotor features of Parkinson's disease (PD). The University of Pennsylvania Smell Identification Test (UPSIT) is one of the most popular tools to assess odor identification. Furthermore, olfactory function has been investigated in monogenic parkinsonism showing different performances in relation to both the gene and type of mutation involved.[6-8] Olfaction is significantly reduced among symptomatic Parkin mutation heterozygotes, whereas it was found to be preserved in affected homozygous and compound heterozygous mutation carriers as in healthy heterozygotes.[6] Indeed, the impaired odor identification detected only in Parkin affected heterozygotes may depend on the fact that these patients are likely to have a typical multifactorial PD, in which the heterozygous Parkin mutation represents only a minor, if not negligible, factor contributing to disease risk. As regards PINK1, odor identification was found to be impaired in both homozygotes and symptomatic heterozygotes, whereas it was entirely normal in healthy heterozygotes.[7] Olfactory disturbances, although less severe than in idiopathic PD (iPD), are also a prominent feature of LRRK2-related PD.[8]

To our knowledge, odor identification performance has not been evaluated in either homozygous or heterozygous SYNJ1 Arg258Gln mutation carriers. Recently, the UPSIT Italian version has been developed and administered to a large number of parkinsonian patients and age-matched healthy controls, and normative values for routine use in Italian population have been provided.[9, 10] Here, we adopted the UPSIT Italian version to explore olfactory function in the Italian subjects with homozygous and heterozygous SYNJ1 Arg258Gln mutations. Furthermore, results were compared with those obtained in iPD patients and healthy controls.

Subjects and Methods


Four groups of subjects were considered for this study: (1) HOM (n = 4): 4 affected subjects homozygous for SYNJ1 c.773G>A mutation from the two Italian families already described[1, 2] (2) HET (n = 4): asymptomatic relatives of HOM patients, heterozygous for SYNJ1 c.773G>A mutation; (3) PD (n = 68): sporadic late-onset Italian iPD; and (4) CTR (n = 61): Italian healthy controls.

PD patients were recruited at the Movement Disorders Unit of University Federico II in Naples, and CTR subjects were recruited among staff and visitors of the same hospital. PD was diagnosed according to the UK Parkinson's Disease Society Brain Bank criteria. For all the four groups of subjects, exclusion criteria were: active upper respiratory tract inflammation; history of diabetes; nose surgery; or head trauma. Moreover, in order to exclude demented subjects, all participants underwent the Mini–Mental State Examination (MMSE) and were excluded if scored below 24. In both HOM and PD patients, motor disability was evaluated with the UPDRS Part III (UPDRS-III).

The study was approved by the local ethics committee, and written informed consent was obtained from all patients according to the Declaration of Helsinki.

Smell Testing

Subjects were tested with a culturally adapted version of the UPSIT for Italian population.[9, 10] The UPSIT is a 40-item, self-administered, multiple- and forced-choice smell identification test. That is, for each odorant, subjects were required to identify the correct answer between four possibilities. The UPSIT consists of 40 standardized encapsulated odors. Raw scores are calculated as the simple number of correct identifications ranging from 0 to 40; higher scores indicate better olfaction. The UPSIT Italian version has been recently implemented; the cultural adaptations used have been described in detail elsewhere, and age- and gender-adjusted normative values have been provided.[8, 9]

Statistical Analysis

Demographics, clinical characteristics, and UPSIT performance were compared among the four groups using univariate analysis of variance and Fisher's tests, as appropriate. Post-hoc analysis was performed with Bonferroni's test. A linear regression model was constructed to assess the association between raw UPSIT score (outcome) and group (HOM, HET, PD, and CTR), adjusting for age, gender, and current smoking status (smoker vs. nonsmoker). In order to evaluate the effect of cognition on the UPSIT performance and considering the relatively small impact of smoking, regression analysis was repeated, substituting current smoking status with MMSE scores. Subsequently, raw UPSIT scores were corrected using normative data for age and gender, and the proportion of subjects with severe olfactory impairment, defined as a score of 18.80 or lower, was compared among the four groups.[10] Computation was supported by SPSS software (version 17.0; SPSS, Inc., Chicago, IL). Significance threshold was set to P < 0.05.


One HOM patient was excluded because MMSE administration was not feasible owing to the severe motor and cognitive disability. Thus, the HOM group included 3 subjects (including HOM3 who scored 21 on the MMSE). Demographic and clinical features of HOM and HET subjects are displayed in Table 1. Clinical information of PD and CTR subjects have been already described in detail elsewhere[9] and are reported in Table S1.

Table 1. Demographic and clinical features of homozygous and heterozygous subjects
  Gender Age Smoking Status Raw MMSE UPDRS-III Disease Duration Raw UPSIT Score SYNJ1 Mutation

1Raw UPSIT scores are shown.

2W, woman; M, man; N, no; Y, yes; Na, not applicable.

HOM1 W 31 N 26 57 3 20 p.Arg258Gln/p.Arg258Gln
HOM2 M 36 N 24 40 8 30 p.Arg258Gln/p.Arg258Gln
HOM3 W 32 N 21 23 6 14 p.Arg258Gln/p.Arg258Gln
HET1 W 70 N 26 Na Na 20 p.Arg258Gln/-
HET2 W 48 N 28 Na Na 30 p.Arg258Gln/-
HET3 W 63 N 30 Na Na 28 p.Arg258Gln/-
HET4 M 67 Y 28 Na Na 25 p.Arg258Gln/-

HOM were significantly younger than CTR, PD, and HET (P < 0.001) and presented more-severe motor symptoms than PD, as measured with UPDRS-III (P < 0.001). Gender distribution and smoking habits were not significant different between groups. When considering UPSIT raw scores, both HET and CTR performed significantly better on the UPSIT than PD (P < 0.001). HOM and PD did not present significantly different UPSIT raw scores. After adjusting for age, gender, and current smoking status, HOM presented significantly worse UPSIT scores, as compared to both HET and CTR, but not as compared to PD (HET = 9.1-point difference, P = 0.029; CTR = 9.8,P = 0.003; PD = 0.7, P = 0.8). In addition, HET showed significant higher UPSIT scores, as compared to PD, but not as compared to CTR, when considering age, gender, and current smoking status as covariates (PD = 8.3, P = 0.002; CTR = 0.7, P = 0.7). When repeating the regression analysis considering the MMSE scores instead of smoking status, HOM presented only a trend toward significance for worse UPSIT scores, as compared to HET (HET = 9.2-point difference, P = 0.066). On the other hand, the other results were largely confirmed. In detail, HOM presented significantly worse UPSIT scores, as compared to CTR, but not as compared to PD (CTR = 9.5, P = 0.015; PD = 1.6, P = 0.7), whereas HET showed significant higher UPSIT scores, as compared to PD, but not as compared to CTR (PD = 7.5,P = 0.005; CTR = 0.2, P = 0.9).

The proportion of subjects with severe olfactory impairment, computed according to previously published normative data, was higher in both HOM and PD (66.6% and 50%, respectively), as compared to both HET and CTR (0% and 6.6%, respectively, P < 0.001; Fig. 1).[10]No significant association was found between severity of olfactory deficit and overall severity of phenotype (data not shown).


Figure 1. Proportion of subjects with severe olfactory impairment in the four groups. Severe olfactory impairment has been defined as a corrected UPSIT score of 18.80 or lower, as already described.[9] Number of subjects for each group is shown in brackets. Whiskers represent the 95% CIconfidence interval for the data above and below the average value. *< 0.001.


Our study demonstrates impaired olfaction in homozygous SYNJ1 Arg258Gln mutation carriers, similar to that observed in PD patients. Our results were largely confirmed when considering cognition as a covariate in the analysis. These findings support olfactory impairment as an additional feature of PARK20. Indeed, olfaction is strongly linked with cognition, and, more in general, with neurodegeneration.[4, 11]Accordingly, a recent study showed that impairment in olfaction predicts future development of cognitive decline and Alzheimer's disease in cognitively intact subjects.[11] Regarding olfaction in monogenic parkinsonism, it has been demonstrated that odor identification is preserved in Parkin homozygotes, whereas it is altered in both PINK1 homozygotes and LRRK2-PD patients.[6-8] Classically, olfactory dysfunction in parkinsonism has been linked with Lewy body (LB) infiltration in both the olfactory bulb and tract. The preserved identification ability reported in Parkin homozygotes, who, despite a large pathological heterogeneity, generally lack LB pathology, would support this association.[12] However, recent evidence shows that olfactory loss may be associated with deposition of pathological aggregates in the rhinencephalon in a variety of neurodegenerative diseases, regardless of the underlying pathological aggregates.[4]Recently, tau-positive neurofibrillary tangles localized predominantly in the SN were reported in a 6-year-old child with a severe seizure disorder and a homozygous truncating mutation within SYNJ1 (c.406C>T, p.R136X). TDP43- and apha-synuclein-positive pathology were not detected. This report expands further the phenotypic spectrum associated with SYNJ1 mutations.[13] However, the truncating mutation present in the child with nigral tauopathy is predicted to cause total loss of SYNJ1 protein activity,[13] whereas the PARK20-causing missense mutation only affects one of the two phosphatase SYNJ1 enzymatic domains, and it might therefore leave substantial residual protein activity intact. Thus, the mechanisms of disease and neuropathology might differ in PARK20 patients. No neuropathology has been reported in adult patients with parkinsonism and SYNJ1 Arg258Gln mutation carriers to date.

SYNJ1 plays a crucial role in the synaptic vesicle-trafficking pathways. There is increasing evidence that defective postendocytic recycling of synaptic vesicles is implicated in the pathogenesis of neurodegenerative diseases. Given that impaired olfaction is considered among the biomarkers of neurodegeneration, our data further support the link between abnormal recycling of synaptic vesicles and neurodegeneration.

The main drawback of our study is the limited sample size precluding definite conclusions about olfaction in SYNJ1-related parkinsonism. However, given that only 6 patients harboring homozygous SYNJ1 mutations have been described thus far, it is an asset to have 3 of them involved in our study. Furthermore, we failed to analyze whether olfactory impairment was perceived by patients at a subjective level.

In conclusion, odor identification is impaired in homozygous SYNJ1 Arg258Gln mutation carriers. These data suggest new insights into the pathophysiology of PARK20-associated parkinsonism.


Leave a Comment

The following required items were not provided or are in the wrong format. Please provide the required responses and submit again:

Comment Title
Comment: 1000 characters
  [[put error message here]]