(XLSX) pone.0198442.s006.xlsx (9.9K) GUID:?1F1E1116-5A1A-469A-8951-D3ADFCAF7559 S7 Table: Comparison with earlier mass spectrometry studies of the perilymph proteome. understand the etiology, pathophysiology and progression of several inner ear disorders. Moreover, the fluid from the inner ear cannot be sampled for micro-chemical analyses from healthy individuals genome (reference list 20972). This was followed by an over-representation test that presents up- or down-regulations of proteins in a selected sample. We choose to present only significant (p 0.05) up- or down-regulated pathways in our results. Statistical methods Analysis was designed to assess the associations between perilymph protein content and clinical parameters, tumor-associated hearing loss and tumor diameter. Only proteins detected in at least 12 out of the 15 samples were included in the final analysis. This cut off was decided upon after analyzing the detection frequency for each protein in the samples, to ensure that there was consistency in terms of missing values. First, we applied univariate linear regression with hearing loss or tumor diameter as dependent variables, and protein levels as explanatory variables. This allowed us to create an overview of the data. Random Forest was used to determine the variable importance of all variables in the data [14] followed by the Boruta method, which is based on repeated Random Forest analyses, to evaluate if the result was independent from random variations [15]. Ethic statements This study was approved by the regional ethics committee in Uppsala (ref 2013/255). Oral and written consent was obtained from all patients prior to surgery and the study complied with the rules of the Declaration of Helsinki. Results Quantification of perilymph proteins In total, MaxQuant identified 314 different proteins; 60 of these proteins were identified in all 15 patients and 130 proteins were only identified once in 15 patients. Ninety-one proteins were detected with a cut-off set to 12 out of 15 patients, and hereafter referred to as frequently identified proteins. In total, 184 proteins were found in four perilymph samples or less (Fig 1). The full list of proteins identified by MaxQuant is provided in supporting information S1 Table. Open in a separate windows Fig 1 The number of proteins recognized in declining order of individuals. This number shows the distribution of 314 proteins recognized in perilymph samples from fifteen individuals. Sixty proteins were recognized in all samples, and 91 in 12 or more samples. Note that 130 proteins were only recognized once in 15 individuals. Mass spectrometer data was analyzed using MaxQuant software. Perilymph proteins and tumor-associated hearing loss Of the 91 frequently-identified proteins N-Desmethyl Clomipramine D3 hydrochloride in the perilymph, four proteins were demonstrated N-Desmethyl Clomipramine D3 hydrochloride by univariate linear regression to be significantly correlated to tumor-associated hearing loss: Ig gamma-4 chain C region (“type”:”entrez-protein”,”attrs”:”text”:”P01857″,”term_id”:”121039″,”term_text”:”P01857″P01857; p = 0.005); Ig kappa chain C region (“type”:”entrez-protein”,”attrs”:”text”:”P01834″,”term_id”:”1160421833″,”term_text”:”P01834″P01834; p = 0.015); Match C3 (“type”:”entrez-protein”,”attrs”:”text”:”P01024″,”term_id”:”119370332″,”term_text”:”P01024″P01024; p = 0.016) and immunoglobulin heavy constant gamma 3 (“type”:”entrez-protein”,”attrs”:”text”:”P01860″,”term_id”:”193806361″,”term_text”:”P01860″P01860; p = 0.023) (Table 2). Table 2 Univariate linear regression results for proteins vs. tumor-associated hearing loss. valuefrom individuals with normal inner hearing function. Therefore, studies of the normal perilymph proteome are not yet possible. Sampling of perilymph is definitely theoretically hard, and several earlier studies possess reported problems with blood contamination. In the present study, careful micropipette aspiration through the RWM, before opening N-Desmethyl Clomipramine D3 hydrochloride the labyrinth, resulted in a low incidence of contaminated samples. The same medical approach, but having a different sampling material and technique, was also reported in another study [19]. Our end result was very beneficial, with only 1 1 out of 16 samples contaminated, compared to 16 FASN out of 24 samples in a earlier report [3]. The total number of recognized proteins in the present study (314) is definitely consistent with earlier reports using non-selective approaches to analyze the proteome of perilymph from individuals with differential inner ear analysis at a group level [4][5][19]. Furthermore, 91 proteins were recognized in 12 to 15 samples, and this was validated by comparison with data arising from earlier MS studies on perilymph [5,19]. Eighty-nine of these regularly recognized proteins were also found in samples from VS individuals in two earlier studies and may become presumed to represent a stable part of the perilymph.