Multiple sclerosis (MS) has long been recognized as a neurological condition that develops due to anomalous function of the adaptive immune system, largely due to the connection of MS to specific human leukocyte antigen (HLA) alleles [1, 2]. Other neurological diseases have also been considered to be impacted by the immune system, but in the absence of clear and significant HLA allele linkages to disease development, the linkage of other neurological diseases to the anomalous function of the adaptive immune system has been limited. However, there is apparently no role for HLA antigen presenting molecules in the function of gamma-delta T-cells, and in the case of Parkinson’s disease (PD), there has been a report of higher levels of gamma-delta T-cells [3]. Also, it has been recently reported that a higher percentage of positive charges, i.e., a higher isoelectric point, among T-cell receptor gamma (TRG) complementarity determining region-3 (CDR3) amino acids (AAs) in Alzheimer’s disease is associated with a higher Braak stage [4], representing more neurofibrillary tangles and a worse clinical situation. Thus, the possibility was considered that specific TRG recombination features could be associated with amyotrophic lateral sclerosis (ALS).

Whole exome sequencing (WXS) blood files representing ALS database of genotypes and phenotypes (dbGaP, phs000101) dataset were mined for the TRG recombination reads, the variable (V)- and joining (J)-gene segments on the reads were identified, and the intervening TRG CDR3 AA sequences were obtained. It is important to note that the algorithm used for the identification of each TRG recombination read requires a validated V and J on one read, thereby identifying the CDR3 to a very high confidence level [5–8]. In general, the application of this algorithm to WXS or RNAseq files does lead to a lower level of recovery of adaptive immune receptor recombination reads than other algorithms, not surprising given the relatively high standard of a verifiable V- and J-gene segments on each sequencing read [8].

The ALS TRG recombinations were compared to TRG recombinations obtained in the same way, using precisely the same algorithm, from other, non-ALS samples (Table 1). Also, the supporting online material (SOM) files (Tables S1–S15) include the complete data collection for the TRG results from mining the genomics files indicated in Table 1 (with the exception of the sequencing read, which is controlled access material). For all comparisons of the ALS TRG recombinations with the TRG recombinations represented by the other datasets, it was clear that there was a very high level of recombinations of the V9-JP gene segments among the ALS cases. The only exception to these disease comparisons and the result of V9-JP specificity for ALS were the TRG recombination reads obtained from the PD blood sample exome files (Table 1).

To further validate the observation of the comparatively high level of TRG V9-JP recombinations from ALS blood exome files, the TRG recombination, sequencing reads were mined from ALS blood RNAseq files (dbGaP, phs002055). The same algorithm was used to mine three RNAseq file datasets, with results indicating that the TRG V9-JP recombination read frequencies from these non-ALS datasets were significantly lower (Table 1).

The data in Table 1, discussed above, suffer from two limitations. First, the recovery of TRG recombination reads from the ALS blood exomes is relatively small. As noted above, these data are sufficient for a relatively high standard of statistical significance (Table 1), but there is no doubt that a much higher level of extraction of recombination reads, inevitable with future datasets with higher fold coverage, would be important for verification of the result reported here. Second, there was no opportunity to compare the recovery of TRG recombination reads from the RNAseq files representing ALS blood samples to the recovery of TRG recombination reads from RNAseq files representing blood samples of other conditions or controls (the non-ALS RNAseq files in Table 1 represent tumor samples). Again, blood to blood sample comparisons will almost certainly be possible in the future due to an increased emphasis on blood biomarkers discoverable via blood-based RNAseq files.

Despite the limitations of the above set of comparisons, it is difficult to ignore a proposal for the development of ALS, and likely other amyloid-based neurological diseases, whereby inflexible cellular membranes, in currently uncertain but possibly disease-specific cell types, facilitate amyloid development [9–11]. Thus, a reduction in, or an immobilization of cell membrane phospholipids could lead to a level of increased cell membrane stiffness that could in turn facilitate amyloid development and cellular dysfunction. And for several decades, there has been published evidence of the beneficial effect of dietary phospholipids in the setting of multiple amyloid-based, neurological diseases [12–16]. This scenario raises the question of whether certain biases in TRG V-J usage, that, for example, favor a juxtaposition of lysines or positive charges, as in the JP gene segment, could lead to the sponging of the negatively charged phospholipids, thereby increasing cell membrane rigidity. Anti-phospholipid antibodies, where positively charged arginines are important for phospholipid binding [17], are causative for anti-phospholipid antibody syndrome [18], a condition that, however, generally does not include symptoms related to ALS or any other amyloid-based, neurological disease, raising questions about the likelihood of TRG phospholipid sponging as having a role in amyloid-based, neurological disease. However, questions could also be raised about the differences in the mechanism or even tissue residency for anti-phospholipid antibodies versus gamma-delta T-cells. This issue of specificity of effect also arises in consideration of other neurological diseases that may be traceable to amyloid development in turn, potentially, indirectly traceable to gamma-delta T-cell sponging of phospholipids. There are no data at all to speak to such specificity of pathological effect as it relates to gamma-delta T-cells, but depending on how future studies may support a role for gamma-delta T-cells, it may be that questions regarding specificity of the pathological effect will be addressed by examining varying gamma-delta T-cell effects on different cell or tissue types or varying effects traceable to subtle differences in gamma-delta receptor-phospholipid affinities.

Finally, regardless of the potential mechanistic role of gamma-delta T-cells, and the JP KKIK AA motif in ALS development, the data reported here raise the question of whether that data can be used for biomarker development to identify persons at risk for ALS or to monitor disease course and severity.