In this study, we quantified that the IP mRNA level, especially the PSMB8 subunit, was expressed significantly more in the PD group than in the HC group. To the best of our knowledge, this is the first study that evaluated IP transcript alteration in the PBMCs of patients with PD. These results are in concordance with those of previous studies that evaluated changes in IP subunits in the central nervous system (CNS) of PD animal models and postmortem PD brain21,22. The present study showed in vivo evidence of IP changes in the peripheral blood of patients with PD.
McNaught et al.10 reported the selective loss of 20S proteasome α-subunits in the substantia nigra pars compacta of PD brain, even though the difference in β-subunit of proteasome, such as IP, was not shown. Recently, Ugras et al. provided more direct evidence regarding IP changes in PD21. They found increased PSMB8 expressions in postmortem brain samples with dementia with Lewy bodies (DLB) and PD compared with HC samples. In the same publication, the proteome analysis revealed the prominent increase of low-molecular-mass protein 7 (LMP7, another common name of PSMB8) in α-Syn-preformed fibril injected mice model and DLB brain21. In PD mice model by rotenone23, PSMB8 and PSMB9 expressions, accompanied with the levels of several antigen presentation-related proteins in the MHC region, were increased in line with the aggregation of α-syn. Moreover, in the α-syn overexpression mice model, insufficient IP assembly led to the accumulation of α-syn24. From these evidences, IP changes appeared clearly in the CNS of both postmortem samples and several mice models of PD. Although no study has evaluated IP in the peripheral blood of patients with PD, a previous study reported a shift in proteasomal enzyme activity, such as ubiquitin in PBMCs25 or in peripheral blood lymphocytes13, using peripheral blood of patients with PD.
Although only the expression of PSMB8 increased in the present study, all three catalytic subunits of IP are often mentioned to simultaneously express after the induction of cytokines such as INF-ɣ through the JAK/STAT pathway26,27. However, even though it was not well known, Barton et al. showed that IP subunits can be expressed independently with IFN-ɣ signaling 28. The different expression levels of IP subunits were found in microgila under normal conditions, and PSMB8 levels in the microglia were not changed even after treatment with IFN-ɣ29. Additionally, when analyzing the proteome of the whole brain and substantia nigra of the PD model, only the increase in PSMB8 was reported, and there was no evidence about other subunits21. Accordingly, we can infer that different catalytic IP subunits can be regulated by distinct inducers or distinct pathways. An antioxidant response element (ARE) consensus sequence exists only in the upstream of the PSMB8 gene and are not present in other IP genes such as PSMB10 or PSMB930. This finding can support the point that IP expression in PD can be regulated by Nrf2, which translocases and binds to the ARE sequence in oxidative stress condition14,27 and leads to an increase of PSMB8 mRNA levels in the PBMCs of patients with PD. Regarding IP assembly, another reasonable explanation appears when PSMB8 is the first requirement for the recruitment and assembly of PSMB9 and PSMB10, respectively31,32. This rule of IP assembly also presents that intermediate IP subtypes exist in some tissues14. In these accepted subtypes, the single intermediate IP (SIP) is only replaced by PSMB8 without PSMB9 and PSMB10 substitutions; thus, this may suggest that SIP can act as one of the main types of UPS in PD. In addition, SIP often appears to be preferred as 20S or 26S form, whereas 20S is mainly responsible for degrading intrinsically disordered protein such as the native α-syn33, which is not degraded by 26S proteasome14,34,35. In addition, SIP expresses capase-like activity same as proteasomes and higher than IP; thus, this catalytic activity can cleave after acidic residues of α-Syn form α-Syn c-truncated species and demonstrated as a pathological form in all synucleinopathies36–38. Given all the above regions, there is a high possibility that SIP exerts detrimental influence on PD progression. Besides, the correlation analysis revealed that the change into a negative correlation of the PSMB10/8 ratio in the drug-treated PD can reflect the possible effect of PD treatment on the PSMB10/8 ratio.
In this study, the PD duration was quite short because this study enrolled first-time clinic visitors. Thus, even the moderate-to-severe PD subgroup defined as H&Y > 2.5 showed similar disease duration with the mild PD group. Moreover, the moderate-to-severe PD group is older than the mild PD group. Thus, their higher H&Y stage can be caused by the aging process than disease progression itself in some cases. Otherwise, atypical parkinsonism can be included in this group because they were firstly diagnosed with PD and their clinical follow-up period was not enough. Thus, this can be a limitation in showing the difference of IP expressions according to disease progression. However, the mild PD or drug-naive PD group showed differences in IP expression more definitely than HC the group. In addition, PSMB10/8 ratio showed significant positive correlation with UPDRS in mild PD group, not moderate-to-severe PD group. This implied that the alteration of IP expression in PD can be an early disease process rather than a result of disease progression. Thus, further evaluation in a large number of patients including those with prodromal PD, advanced PD and atypical parkinsonism would be informative.
In the recent review15 by Kimura et al., although IP has three inducible subunits including PSMB8-10, PSMB8 showed its roles in many diseases distinctively, whereas other subunits did not contribute. Indeed, PSMB8 has demonstrated its involvement in autoimmune disease, lipodystrophy, lipid metabolic disorders, and diabetes mellitus15. In addition, PSMB8 was reported to be responsible for cell growth and proliferation activities39. Therefore, the common IP inhibitor ONX-0914, used for treating some conditions such as inflammatory diseases, is the selective LMP7 (PSMB8) inhibitor15,40. With these supporting findings along with our data of the unique increased PSMB8 mRNA level, the novel unknown mechanism of IP in PD can pave possible approaches for PD treatment.
Previous studies evaluating standard proteasome activity in PD showed the decreased function of UPS system in the brain and peripheral blood of PD11,13,25. However, the current experiment showed that one subtype of proteasome complex, IP, especially a SIP with PSMB8 expressed more in PD compared to HC. In especial, this is more significant in mild or drug-naive PD. This implied that, during the disease process of PD, decreased standard proteasomal activity can be a resultant condition than an early initial cause such as exceptional genetic mutation. Increased IP expression in the current study can be a compensating mechanism against to progressive α-Syn accumulation or immune dysfunction. And it is possible that early studies did not represent specific activation of IP in PD process. Otherwise, the increased expression of PSMB8 could be a compensatory mechanism for dysfunction or incomplete assembly of IP complex. Although why IP is involved in PD mechanism is unclear, it can be presumed as followings. First, IP catalytic subunit is more chymotrypsin-prone than the standard proteasome. This feature promotes the degradation of hydrophobic proteins like α-Syn oligomer or aggregates21. Then the degraded α-Syn residue by IP might be presented as antigen by MHC class I or II41. Otherwise, IP activation itself could be the result not by abnormal α-Syn aggregates, but by immune activation observed in PD3.
In conclusion, to the best of our knowledge, this is the first evidence about the altered expression of IP PSMB8 subunit in the PBMCs of patients with PD. Further studies including advanced PD and prodromal PD would be informative to understand the mechanism of PD regarding IP.