Alpha-Synuclein Pathology in the Submandibular Gland of LRRK2 p.G2019S Mutation Carriers.

Background. The presence of intraneuronal aggregates of phosphorylated alpha-synuclein (pAS), the histological hallmark of Parkinson disease (PD), has been already demonstrated to be present in the autonomic nerve bres that innervate the submandibular gland in approximately 75% of living PD patients. The presence of pAS in the peripheral autonomic nervous system in carriers of LRRK2 mutations has not been studied so far. The objective of the current study is to evaluate the presence of abnormal pAS aggregates in the submandibular gland tissue of LRRK2 p.G2019S mutations carriers. Methods. This is a prospective observational study conducted between 2014 and 2015 at Hospital Clínic de Barcelona, Spain. A random sample of nine asymptomatic LRRK2 (aLRRK2) and 11 LRRK2 associated PD (LRRK2-PD) patients were recruited among a cohort of LRRK2-PD patients and their relatives already identied at our centre. All the participants underwent transcutaneous needle core biopsy of the submandibular gland under ultrasound guidance. The presence of pAS was assessed in all the participants by immunohistochemistry using anti-Serine 129-phosphorylated AS antibody. Results. Submandibular biopsy material containing glandular parenchyma was obtained in 4 (44.44%) aLRRK2 and in 6 (54.55%) LRRK2-PD patients. Aggregates of pAS were detected in the glandular parenchyma in one of the four (25%) aLRRK2 subjects and in none (0%) of the LRRK2-PD patients. Conclusions. Our study shows that pAS aggregates obtained by needle core biopsy of the submandibular gland are infrequent in LRRK2 mutation carriers but may be detected in asymptomatic mutation carriers. The low rate of pAS positive biopsies suggests either a different physiopathology between LRRK2-related and idiopathic PD or that a one-time unilateral submandibular gland biopsy is not the optimal procedure for the study of synuclein aggregation in LRRK2 mutation carriers. and controls. Results from the iPD and the from the aLRRK2 and LRRK2-PD participants.

gland are infrequent in LRRK2 mutation carriers but may be detected in asymptomatic mutation carriers. The low rate of pAS positive biopsies suggests either a different physiopathology between LRRK2-related and idiopathic PD or that a one-time unilateral submandibular gland biopsy is not the optimal procedure for the study of synuclein aggregation in LRRK2 mutation carriers.

Background
Abnormal aggregates of phosphorylated alpha-synuclein (pAS) are the major component of the intraneuronal inclusions known as Lewy bodies and Lewy neurites, the histological hallmark of Parkinson disease (PD) 1 . Recently, pAS deposits were detected in the submandibular gland in living patients with idiopathic PD 2,3 , and also in the prodromal phase of this condition 4 , providing biological evidence of ongoing disease at its earliest stage.
We have assessed the presence of pAS aggregates in the submandibular gland of LRRK2 p.G2019S mutations carriers, both in those with manifest PD (LRRK2-PD) and in asymptomatic carriers (aLRRK2).
As LRRK2-related CNS neuropathology is not necessarily associated with pAS aggregates 5 , the detection of pAS in peripheral biopsies in LRRK2-PD cases would identify those with an underlying Lewy type pathology which could be important for implementing potential anti-synuclein therapies. In aLRRK2, the presence of peripheral pAS could re ect an underlying synucleinopathy and constitute a risk marker for the later development of manifest PD.

Methods
This is a prospective observational study conducted between July 2014 and May 2015 at the Hospital Clinic de Barcelona, Spain. The ethical committee at our institution approved the study and all participants gave their written informed consent.

Participants
At our institution ve groups of subjects underwent biopsy of the submandibular gland for research: aLRRK2, LRRK2-PD patients, idiopathic (iPD), idiopathic REM sleep behaviour disorder (IRBD) and controls. Results from the IRBD, iPD patients and controls were previously reported 4 . Herein we present the results from the aLRRK2 and LRRK2-PD participants.
Asymptomatic LRRK2 mutation carriers and LRRK2-PD patients. Nine aLRRK2 carriers of the p.G2019S mutation and eleven LRRK2-PD patients were selected among a cohort of LRRK2-PD patients and their relatives. Participants had been already identi ed at our centre, had been screened for LRRK2 mutations as previously described 6 , and were positive for the LRRK2 p.G2019S mutation. Participation in the study was proposed to all who were alive and accessible. PD was diagnosed according to United Kingdom PD Society criteria with the exception that a positive family history was not considered an exclusion criterion 7 .
Exclusion criteria in all participants were current or past medical history of disorders affecting the salivary glands (e.g., chronic sialadenitis, abscesses, active neoplasms, Sjögren syndrome), coagulation disorders, and anticoagulant or antiplatelet drug intake. Echogenicity of the substantia nigra was assessed by means of transcranial sonography as previously described. 15 Dopamine transporter imaging with 123 I-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3uoropropyl)-nortropane single photon emission computed tomography (DaT-SPECT) was performed as previously described. 16 Procedures Submandibular gland biopsy and alpha-synuclein immunohistochemistry of the submandibular gland were assessed as described previously 4 . Brie y, a core needle biopsy was performed in all participants using a commercial 16-gauge needle. Biopsies were performed unilaterally under ultrasound guidance after subcutaneous in ltration of local anesthesia. Between 2-4 punctures were performed by an experienced radiologist who was masked to the clinical status of the participants. After each procedure, specimens were immediately processed and serial 4 µm sectioning of the whole specimens was performed. Three slides were selected from the rst, middle or nal third of the blank slide set from each subject and were stained with haematoxylin-eosin. Immunohistochemistry was performed on the third set with most submandibular parenchyma and/or nervous s tissue using an automated Dako Autostainer using anti-Serine 129-phosphorylated alpha-synuclein antibody (Wako clone pSyn#64;, Pure Chemical Industries, Osaka, Japan)), as described previously 4 . pAS immunoreactivity in nervous structures within or outside the submandibular gland parenchyma was assessed as present or absent. At the time of the histopathologist examination, neuropathologists were blinded regarding the clinical condition of the individual's tissues.

Results
Demographic and clinical data of participants are summarized in Tables 1 and 2. Biopsy specimens contained submandibular gland parenchyma in 6 of the 11 (54.55%) LRRK2-PD patients and in 4 of the 9 (44.44%) aLRRK2. The remaining samples contained periglandular connective tissue with variable amount of vessels and nerve bres or muscle.  Aggregates of pAS were detected in the glandular parenchyma in one of the four (25%) aLRRK2 subjects ( Fig. 1) and in none (0%) of the six LRRK2-PD patients with available glandular tissue. pAS aggregates in the aLRRK2 individual were identi ed in nerve structures of the connective tissue within the gland, rarely surrounding individual glands. This type of aggregates has been shown to co-distribute with tyrosinehydroxylase positive sympatethic nerve bres 4 .
None of the aLRRK2 or the LRRK2-PD showed extraglandular pAS aggregates.
Asymptomatic LRRK2 p.G2019S carriers (Tables 1 and 2) The mean age of the aLRRK2 was 48.89 ± 8.43 years. The mean UPSIT score was 33.56 ± 2.19 points, and none of the nine aLRRK2 had hyposmia. None of the aLRRK2 had depression, constipation or RBD by the RBSQ. Seven of the nine aLRRK2 underwent a DaT-SPECT, which was normal in all of them. Three out of seven (42.9%) individuals who underwent transcranial sonography had hyperechogenicity of the substantia nigra .
The asymptomatic LRRK2 mutation carrier who showed pAS pathology in the submandibular glandular parenchyma was a 60-year-old woman. Her score in the NMSQ was 14, with all points got from the sleep subscore due to insomnia. Her neurological examination was normal, without evidence of parkinsonian signs. A conducted interview detected no history of dream-enacting behaviours, constipation or depression. She had no hyposmia. The echogenicity of the substantia nigra and the DaT-SPECT were normal. At the age of 65, she is still asymptomatic.

Discussion
To the best of our knowledge this is the rst study that assessed the presence of pAS pathology in the submandibular gland of LRRK2 mutation carriers. pAS aggregates were found in one of the four aLRRK2 in whom glandular parenchyma was available after transcutaneous needle core biopsy. None of the LRRK2-PD patients had pAS aggregates in the submandibular gland or extraglandular tissues. Submandibular gland parenchyma was obtained only in around 50% of the LRRK2-PD and aLRRK2 individuals.
The absence of pAS positivity in manifest LRRK2-PD was not expected, since pAS accumulation in the peripheral autonomic nervous system is thought to re ect an ongoing synucleinopathy, which occurs in the majority of, but not all, LRRK2-G2019S-PD cases. Still, in a study of alpha-synuclein aggregation in cerebrospinal uid by real-time quaking-induced conversion (RT-QuIC) 17 the percentage of positive cases in LRRK2-PD was also much lower than in iPD (40% vs 90%). The pleiomorphic pathology of LRRK2-PD linked to the p.G2019S mutation may in part explain the results. Other possible explanations include the possibility that, unlike in iPD, peripheral autonomic nervous system Lewy type pathology is less prominent or even absent in p.G2019S LRRK2-PD patients. This could be supported by the notion that dysautonomia seems to occur less frequently in p.G2019S LRRK2-PD than in iPD 18 , although no de nitive data are available. Also possible is that synuclein pathology might have been present in the peripheral autonomic system at an early disease stage and migrated later centripetally to the central nervous system, becoming undetectable in manifest LRRK2-PD, as has been speculated for idiopathic PD 19 . Finally, a sampling bias with under-representative tissue samples cannot be excluded.
The number of submandibular glands with pAS positivity among aLRRK2 was low (25%). Of interest, the percentage of positive cases is close to the proportion of aLRRK2 reported to have misfolded synuclein in the cerebrospinal uid (18.8%) assessed by RT-QuIC 17 . Factors such the relatively young age of the patients may also in uence these results since parkinsonism appears generally late in LRRK2-PD. However, even if our patients possibly destined to develop PD might have been disease free at the time of the biopsy, manifest LRRK2-PD were also pAS negative.
The main limitations of our study are the small sample size which precludes generalization of the results and the low frequency of glandular tissue obtained with unilateral transcutaneous needle biopsy of the submandibular gland despite the use of ultrasound guiadance. In previous studies in living PD using a similar biopsy methodology but without ultrasound guidance, submandibular glandular tissue was not obtained in 20-24% of participants 2-3 . Bilateral transcutaneous needle biopsies of the submandibular gland in PD patients seemed recently feasible and safe, showing a better tissue acquisition 21 . Technical re nement of the procedure is needed to improve the ability to obtain glandular parenchyma in living subjects.
In conclusion, our study shows that pAS aggregates obtained by needle core biopsy of the submandibular gland may be detected in aLRRK2, suggesting that Lewy type pathology is already ongoing in a subset of subjects. However, the low rate of submandibular gland tissue obtained by the biopsies and the fact that none of the LRRK2-PD patients showed pAS aggregates suggest that a onetime unilateral submandibular gland biopsy may not be the optimal procedure for the study of pAS in LRRK2 p.G2019S mutation carriers. Other peripheral tissues such as skin or minor salivary glands that are easily accessible may prove a better target for tissue-based studies of synuclein aggregates in LRRK2 disease. Yet emerging synuclein ampli cation methods in cerebrospinal uid 16  Ethics approval and consent to participate: The ethical committee at Hospital Clínic de Barcelona approved the study and all participants gave their written informed consent. Figure 1 A: Representative histological image of the submandibular biopsy (haematoxylin-eosin stain). B:

Figures
Immunohistochemistry for phosphorylated alpha-synuclein shows small aggregates surrounding glandular structures (brown signal, arrows) in areas corresponding to the course of autonomic nerve bres. Scale bars: 20 μm.